WO2022115774A1 - Systèmes et procédés pour une attribution et une optimisation dynamiques de bande passante - Google Patents

Systèmes et procédés pour une attribution et une optimisation dynamiques de bande passante Download PDF

Info

Publication number
WO2022115774A1
WO2022115774A1 PCT/US2021/061176 US2021061176W WO2022115774A1 WO 2022115774 A1 WO2022115774 A1 WO 2022115774A1 US 2021061176 W US2021061176 W US 2021061176W WO 2022115774 A1 WO2022115774 A1 WO 2022115774A1
Authority
WO
WIPO (PCT)
Prior art keywords
bandwidth
computing device
bandwidth consumption
loln
threshold
Prior art date
Application number
PCT/US2021/061176
Other languages
English (en)
Inventor
Michael Paul Lovett
Kevin David Williams
David Frederick Giannini
Original Assignee
Gatespeed, Inc.
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Gatespeed, Inc. filed Critical Gatespeed, Inc.
Publication of WO2022115774A1 publication Critical patent/WO2022115774A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0876—Network utilisation, e.g. volume of load or congestion level
    • H04L43/0882—Utilisation of link capacity
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/16—Threshold monitoring
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/76—Admission control; Resource allocation using dynamic resource allocation, e.g. in-call renegotiation requested by the user or requested by the network in response to changing network conditions

Definitions

  • TCP Transmission Control Protocol
  • IP Internet Protocol
  • Some known data networks tend to treat Internet Protocol (IP) traffic equally on a first- come, first-served basis, even though traffic generated by different applications may have different characteristics. Users have different expectations for speed, latency and reliability depending on the application in use. For example, users of a compute device downloading a compact disk or other disk image such as ISO image, uploading a website for deployment or engaging in peer-to- peer file sharing expect a substantial time period to accomplish the data transfer. On the other hand, users of a compute device checking e-mail or viewing websites transfer data in short, relatively infrequent bursts, and expect a more interactive experience. Because the data transferred by IP is given equal priority, users running applications characterized by short bursts of interactive traffic are adversely affected by applications characterized by bulk transfers of non-interactive data.
  • IP Internet Protocol
  • bandwidth management tools allow for the arbitrary restriction of bandwidth consumption imposed upon each user. This solution merely inhibits individual users to an assigned bandwidth limit. This static solution does not adapt bandwidth allocation per users’ consumption requirements at a given time relative to other users’ consumption. In other words, each user suffers restricted bandwidth, regardless of current network capacity or relative priority among users and applications.
  • Methods and systems are described herein for dynamic bandwidth allocation and optimization to address one or more of the problems mentioned above.
  • the methods and systems described herein may operate in numerous hardware, software, and/or network environments. That is the methods and systems provide improved performance whether on physical or virtual hardware.
  • the methods and systems may be implemented as code that interacts with an operating system and manipulates the queuing and client-service functions of the operating system.
  • the methods and systems may be implemented using one or more cloud components/services and/or features of an integrated circuit.
  • the methods and systems may use an abstraction layer that allows for implementation on different chips and systems. It should be noted that any of the embodiments below may be performed in each of these environments and/or a combination of these environments.
  • the methods and systems described herein provide an environment for dynamic bandwidth allocation and optimization that accelerates throughput of packets.
  • the methods and systems described herein provide such efficiencies that the methods and systems observe, record, and/or process data with a higher granularity.
  • the methods and system may poll data in nanosecond increments, which allows for near real-time or real-time polling. Accordingly, the methods and systems do not need to rely on sampling data and may instead generate actual byte counts thus replacing estimations with actual counts.
  • DPDK Data Plane Development Kit
  • the methods and systems may enforce different DPDK rules dynamically.
  • the methods and systems may implement nanosecond policing activity.
  • the methods and systems may monitor network traffic for compliance with rules and enforce those rules (e.g., by discarded data immediately, marking data as non-compliant, etc.).
  • the methods and systems may take action in real-time (e.g., in terms of nanoseconds and/or microseconds).
  • the efficiencies and higher granularity allow the methods and systems to analyze, systematically extract information from, and/or otherwise deal with data sets that are too large or complex to be dealt with by traditional data-processing application software (e.g., Big Data).
  • traditional data-processing application software e.g., Big Data
  • the higher granularity provides additional benefits for artificial intelligence and machine learning applications.
  • the methods and system may now use the Big Data database and may now characterize actual data (i.e., not sampled data). Accordingly, the methods and system may now train model and perform analyses on larger data sets and/or in near real-time or real-time.
  • the methods and systems may use one or more features that may work separately or in conjunction.
  • the system may use express path forwarding as opposed to TCP/IP stack.
  • conventional systems may use a TCP/IP stack for communications.
  • the methods and systems may optimize packet flow (e.g., from an operating system to the Internet) by establishing an express path as opposed to managing and polling the different intercommunications (e.g., to determine if there is bandwidth).
  • the system may also use IP filtering as opposed to IP tables.
  • IP tables For example, conventional systems may use IP tables to manage network traffic.
  • the methods and systems may use dynamic filters to automatically move packets through the express path.
  • the system may also use improve proxy.
  • conventional systems may use a proxy server to manage network traffic.
  • the methods and systems may add a proxy to the software, which may provide acceleration to a CPU through the operating system. Using the proxy, the system may terminate connections and issue a different type of congestion control.
  • the system may also use a scheduler for dynamic adjustment of packet size. For example, while conventional systems may use a scheduler with multi-core systems, the methods and systems may implement a more efficient scheduler, which may dynamically adjust the packet size as the packets sizes are processed on any hardware chip, processor, etc.
  • an amount of aggregated bandwidth consumption for a set of computing devices on a network may be determined for a first time period. An amount of available bandwidth on the network may be determined for the first time period.
  • a ratio of the amount of aggregated bandwidth consumption to the amount of available bandwidth may be determined.
  • a bandwidth threshold may be determined based on the ratio.
  • a rate limit for a first computing device (of the set of computing devices) may be dynamically set based on a comparison of the bandwidth threshold to bandwidth consumption of the first computing device.
  • an amount of aggregated bandwidth consumption for a set of computing devices on a network may be determined.
  • An amount of available bandwidth on the network may be determined.
  • a ratio of the amount of aggregated bandwidth consumption to the amount of available bandwidth may be determined.
  • a bandwidth threshold may be determined based on the ratio or preset by the system.
  • First bandwidth consumption of a first computing device (of the set of computing devices) that is measured for a first time period may be determined.
  • Second bandwidth consumption of the first computing device that is measured for a second time period (of different length than the first time period) may be determined.
  • a rate limit for the first computing device may be dynamically set based on a comparison of the bandwidth threshold to at least one of the first bandwidth consumption or the second bandwidth consumption or preset by appropriate algorithms learned based on Big Data analyses and calibration for the specific bandwidth patterns
  • an amount of aggregated bandwidth consumption for a set of computing devices on a network may be determined for a first time period.
  • An amount of available bandwidth on the network may be determined for the first time period.
  • a ratio of the amount of aggregated bandwidth consumption to the amount of available bandwidth may be determined.
  • a bandwidth threshold may be determined based on the ratio.
  • a first computing device of the set of computing devices may be classified based on a comparison of the bandwidth threshold to bandwidth consumption of the first computing device.
  • a rate limit for a first computing device of the set of computing devices may be dynamically set based on the classification of the first computing device.
  • FIG. l is a schematic block diagram of a computer system in which dynamic bandwidth allocation can be performed, according to an embodiment.
  • FIG. 2 is a schematic illustration of a bandwidth allocation platform, according to an embodiment.
  • FIGS. 3-5 are flowcharts of processes for providing dynamic bandwidth, according to various embodiments.
  • FIG. 6 is an exemplary diagram of dynamic bandwidth allocation, according to an embodiment.
  • FIG. 7 is an exemplary diagram of a system for a dynamic bandwidth allocation platform, according to an embodiment.
  • FIG. 8 is an exemplary diagram of a software stack architecture for a dynamic bandwidth allocation platform, according to an embodiment.
  • FIG. 9 is an exemplary diagram of a LAN-to-WAN arrangement for a dynamic bandwidth allocation platform, according to an embodiment.
  • FIG. 10 is a table comparison of the benefits of the dynamic bandwidth allocation platform, according to an embodiment.
  • the methods and systems described herein may operate in numerous hardware, software, and/or network environments. That is the methods and systems provide improved performance whether on physical or virtual hardware.
  • the methods and systems may be implemented as code that interacts with an operating system and manipulates the queuing and client-service functions of the operating system.
  • the methods and systems may be implemented using one or more cloud components/services and/or features of an integrated circuit.
  • the methods and systems may use an abstraction layer that allows for implementation on different chips and systems. It should be noted that any of the embodiments below may be performed in each of these environments and/or a combination of these environments.
  • the system described herein may comprise software that accelerates, optimizes, and/or dynamically manages performance (e.g., in LAN-to-WAN arrangements) resulting in faster packet forwarding and Internet connectivity.
  • the system may be used to secure Work-From-Home (“W-F- H”) VPN connections to/from an office headquarter, branch, and/or datacenter location. Additionally or alternatively, the system may be used to for shared public and private Wi-Fi in offices, apartment buildings, hotel/motels, transportation systems, and/or distance learning for universities and schools. Additionally or alternatively, the system may be used to for video distribution, videoconferencing, and/or live streaming to/from the public cloud/edge. For example, over the last five years W-F-H activity grew forty-four percent and over the previous ten years it grew ninety-one percent. The top needs for the W-F-H activity is reliability, faster broadband Internet, greater Wi-Fi speed/range, connectivity for PC/laptop devices, increased security, more reliable videoconferencing services, Landline/VoIP support, and/or faster mobile access.
  • W-F- H Work-From-Home
  • the system may be used to for shared public and private Wi-Fi in offices, apartment buildings, hotel/motels, transportation systems, and/or distance
  • the system may use a software stack architecture for a dynamic bandwidth allocation platform as shown in FIG. 8.
  • architecture 800 may provide dynamic bandwidth allocation for network optimization as discussed below).
  • Architecture 800 may also provide integration of dynamic TCP congestion control algorithms such as the proxy for edge “bump-in-the wire” and continuous self-learning (as discussed below).
  • a proxy that provides congestion control at the edge provides more efficient traffic throughput during peak volumes from LAN-to-WAN. Accordingly, the system experiences close to 100% peak circuit utilization, less packet loss (which results in less re-transmissions) and avoids ISP “Buffer Bloat” (which results in less latency and jitter).
  • variable traffic flows protects voice, VPN & burstable traffic sessions, shields casual users from Data Hogs (e.g., high data users), optimizes heavy users’ packet flows (e.g., as customary in video applications), and ensures that traffic can be prioritized by application, subnets, users, and specific device groups.
  • dynamic bandwidth assessment allows for dynamic window sizing and better efficiencies for reducing congestions.
  • the system may use DPDK embodiments, and the software in such embodiments may enable faster inter-stack and CPU communications through the use of express path forwarding engine, nanosecond clocking, and/or drivers/OS stack improvements (as discussed below).
  • a governor e.g., governor 45 (FIG. 9) may determine how to forward packets through the system.
  • the governor may monitor and manage the LAN-to-WAN traffic in real time to ensure data flows efficiently to/from the WAN connection.
  • the system may use dynamic filters to automatically adjust the flow of the packets through an express path platform, which eliminates the need for hardcoded tables.
  • the benefits of the dynamic bandwidth allocation platform allow for virtually the same throughput with one fewer active core, thus lowering power consumption per GBPS of throughput by approximately forty-percent. Accordingly, the system achieves more efficiency and speed, which results in less power and processing cost.
  • an apparatus includes a threshold calculation module implemented in at least one of a memory or a processing device, a use identification module operatively coupled to the threshold calculation module, and a rate limiting module operatively coupled to the use identification module.
  • the threshold calculation module is configured to receive at a time a datum associated with use of a network by a set of compute devices using the network.
  • the threshold calculation module is also configured to calculate a threshold value based, at least in part, on the datum and a bandwidth capacity of the network at the time.
  • the use identification module is configured to receive the threshold value from the threshold calculation module.
  • the use identification module is also configured to receive a usage value associated with bandwidth use of a compute device from the set of compute devices and associated with a time period immediately preceding the time.
  • the use identification module is further configured to compare the usage value with the threshold value to determine a usage category associated with the compute device.
  • the rate limiting module is configured to receive an indication of the usage category from the use identification module.
  • the rate limiting module is also configured to limit bandwidth use of the network for the compute device based on the indication.
  • a non-transitory processor-readable medium stores code that represents instructions to be executed by a processor.
  • the code includes code to cause the processor to receive, at a time, a signal associated with use of a network by a compute device associated with a user.
  • the code also includes code to cause the processor to calculate a first usage value based, at least in part, on the signal associated with use of the network.
  • the first usage value is associated with bandwidth use (1) of the user for a first time period immediately preceding the time and (2) associated with the network.
  • the code further includes code to cause the processor to calculate a second usage value based, at least in part, on the signal associated with use of the network.
  • the second usage value is associated with bandwidth use (1) of the user for a second time period immediately preceding the time and (2) associated with the network.
  • the first time period is different from the second time period.
  • the code further includes code to cause the processor to compare the first usage value and the second usage value with a value of a threshold associated with usage of the network at the time to determine a usage category associated with the user.
  • the code also includes code to cause the processor to send a signal to limit bandwidth use of the user in the network for a time period after the time based, at least in part, on the usage category.
  • a non-transitory processor-readable medium stores code that represents instructions to be executed by a processor.
  • the code includes code to cause the processor to receive, at a first time, a first datum associated with use of a network by a set of compute devices using the network.
  • the code also includes code to cause the processor to calculate a first threshold value based, at least in part, on the first datum and a bandwidth capacity of the network at the first time.
  • the code further includes code to cause the processor to receive a first usage value associated with bandwidth use of the network of a compute device from the set of compute devices for a time period immediately preceding the first time.
  • the code further includes code to cause the processor to send, at a second time after the first time, a signal to limit bandwidth use in the network of the compute device based, at least in part, on the first threshold and the first usage value.
  • the code also includes code to cause the processor to receive, at a third time after the second time, a second datum associated with use of the network by the set of compute devices using the network.
  • the code also includes code to cause the processor to calculate a second threshold value based, at least in part, on the second datum and a bandwidth capacity of the network at the third time. The second threshold is different from the first threshold.
  • the code further includes code to cause the processor to receive a second usage value associated with bandwidth use of the network of the compute device for a time period immediately preceding the third time. The second usage value is substantially equal to the first usage value. Because the threshold has decreased, the bandwidth use of the compute device for the time period immediately preceding the third time is no longer deemed to be excessive and/or offending.
  • the code also includes code to cause the processor to send, at a fourth time after the third time, a signal to not limit bandwidth use in the network of the compute device based, at least in part, on the second threshold and the second usage value.
  • a compute device is intended to mean a single compute device or a combination of compute devices (e.g., mobile devices or IoT sensor devices with access to a certain network, etc.).
  • FIG. l is a schematic block diagram of a computer system in which dynamic bandwidth allocation can be performed, according to an embodiment.
  • the computer network system 100 includes at least one compute device 101 a- 101 n, a bandwidth management platform 103, and at least one gateway device 109, which can be operatively coupled to one or more compute devices lOla-lOln or other gateway devices 109 via a communication network 105.
  • the bandwidth management platform 103 is shown in FIG. 1 as being within the gateway device 109, in other embodiments, the bandwidth management platform (or a portion thereof) can be operatively coupled to the gateway device 109 via the communication network 105 and/or the Local Area Network 107.
  • any of the devices or platforms of the computer network system 100 can be equipped with local memory/storage spaces (shown as memory 121a-121n and memory 125 in FIG. 1). Furthermore, the devices and platforms of the system 100 may have access to centralized or distributed memory/storage spaces (not shown in FIG. 1) through the communication network 105. Additionally, a compute device lOla-lOln and a gateway device 109 each can include one or more processors, performing processes associated with the services provided to the compute devices lOla-lOln (shown as processors 119a-l 19n and processor 123 in FIG. 1). For example, the bandwidth management platform 103 can be stored in the memory 125 and executed at the processor 123, as described in further detail herein. Thus, FIG. 1 is merely an example illustrating the types of devices and platforms that can be included within a computer network system 100.
  • one or more components (or the functions performed thereby) of computer network system 100 may be performed by an abstraction layer, cloud components, and/or an integrated circuit.
  • computer network system 100 may use an abstraction layer that allows for implemented on different chips and systems.
  • the abstraction layer may include a set of functions, written according to the abstraction layer specification, that provides interfaces between an operating system and platform transport mechanism.
  • the set of functions may be generic such that bandwidth management platform 103 may be implemented on different chips and systems.
  • computer network system 100 may be implemented in cloud components (e.g., as described in FIG. 7 below).
  • computer network system lOO may access a set of functions/services available from cloud components.
  • one or more components of computer network system 100 may be implemented in an integrated circuit.
  • the gateway device 109 can be in communication with any number of compute devices lOla-lOln via a Local Area Network (LAN) 107.
  • the gateway device 109 can further be configured to communicate with a wide area network (WAN) or the Internet (shown as the communication network 105).
  • the gateway device 109 can facilitate communication between the compute devices lOla-lOln and/or any other compute device with LAN 107, with any compute device 130 (e.g., personal computers, mobile devices, web servers, file servers and/or the like) operatively coupled to the gateway 109 via the communication network 105.
  • the bandwidth management platform 103 can allocate bandwidth to the compute devices lOla-lOln via the gateway device 109, based on bandwidth characteristics further described herein with respect to FIG. 2.
  • Communication network 105 can be any communication network, such as the Internet, operatively and communicatively coupling the compute device lOla-lOln and/or the gateway device 109 with another compute device 130 (e.g., personal computers, mobile devices, web servers, file servers, IoT sensors, and/or the like).
  • Communication network 105 can be any network or combination of networks capable of transmitting information (e.g., data and/or signals) and can include, for example, a telephone network, an Ethernet network, a fiber-optic network, a wireless network, a satellite, a cellular network and/or the like.
  • communication network 105 and/or LAN 107 can include multiple networks operatively coupled to one another by, for example, network bridges, routers, switches and/or gateways.
  • the compute device lOla-lOln can be operatively coupled to a cellular network; and the gateway device 109 and/or the bandwidth management platform 103 can be operatively coupled to a fiber-optic network.
  • the cellular network and fiber-optic network can each be operatively coupled to one another via one or more network bridges, routers, switches, and/or gateways such that the cellular network and the fiber-optic network are operatively coupled to form a communication network.
  • the cellular network and fiber-optic network can each be operatively coupled to one another via one or more additional networks.
  • the cellular network and the fiber-optic network can each be operatively coupled to the Internet such that the cellular network, the fiber-optic network and the Internet are operatively coupled to form a communication network.
  • the compute device lOla-lOln is operatively coupled to communication network 105 via LAN 107 and network connection(s) 111 and 113; gateway device 109 is operatively coupled to communication network 105 via network connection(s) 117; and the compute device 130 is operatively coupled to communication network 105 via network connection(s) 115.
  • Network connections 111, 113, 115, and 117 can be any appropriate network connection for operatively coupling compute device lOla-lOln, gateway device 109, and the compute device 130.
  • a network connection can be a wireless network connection such as, for example, a wireless fidelity (“Wi-Fi”) or wireless local area network (“WLAN”) connection, a wireless or satellite wide area network (“WWAN”) connection, a Virtual Private Network (VPN), and/or a cellular connection.
  • Wi-Fi wireless fidelity
  • WLAN wireless local area network
  • WWAN wireless or satellite wide area network
  • VPN Virtual Private Network
  • a network connection can be a wired connection such as, for example, an Ethernet connection, a digital subscriber line (“DSL”) connection, a broadband coaxial connection, powerline and/or a fiber-optic connection.
  • DSL digital subscriber line
  • a computer network system 100 can include more than one compute device lOla-lOln, more than one bandwidth management platforms 103, and more than one gateway device 109.
  • a compute device lOla-lOln, a bandwidth management platform 103, a gateway device 109, and/or compute device 130 can be operatively coupled to LAN 107 and/or the communication network 105 by heterogeneous network connections.
  • a first compute device lOla-lOln can be operatively coupled to LAN 107 and/or the communication network 105 by a WWAN network connection
  • another compute device 101a- 101h can be operatively coupled to LAN 107 and/or the communication network 105 by a DSL network connection
  • a compute device 130 can be operatively coupled to the communication network 105 by a fiber-optic network connection.
  • the gateway device 109 can be, for example, an access point configured to provide access to the communication network 105 for the compute device lOla-lOln via LAN 107.
  • the gateway device 109 can include a set of modules (e.g., the bandwidth management platform 103) configured to perform a set of tasks.
  • the modules can be software modules (e.g., stored in memory and executed by a processor), hardware modules, and/or any combination thereof.
  • Such modules can include, for example, the bandwidth management platform 103 and its associated modules (as shown in FIG. 2), a module used to convert a data packet from a format of the LAN 107 to a format of the communication network 105 and vice versa, and/or the like.
  • the gateway device 109 includes a module used to convert a data packet from a LAN 107 to a format of the communication network 105 (and vice versa)
  • the gateway device 109 can have a first interface operatively coupled to the LAN 107 using a first network medium and/or employing a first message transport type or protocol and a second interface operatively coupled to the communication network 105 using a second network medium and/or employing a second message transport type or protocol.
  • the LAN 107 and the gateway device 109 can use a common network medium and/or message transport type.
  • bandwidth management platform 103 and/or one or more modules may be implemented in a cloud environment and/or an integrated circuit.
  • bandwidth management platform 103 may be implemented using DPDK, which provides a set of data plane libraries and network interface controller polling-mode drivers for offloading TCP packet processing from the operating system kernel to processes running in user space.
  • DPDK provides a set of data plane libraries and network interface controller polling-mode drivers for offloading TCP packet processing from the operating system kernel to processes running in user space.
  • Bandwidth management platform 103 may use this offloading to achieve higher computing efficiency and higher packet throughput than that which is possible using the interrupt- driven processing provided in the kernel.
  • Bandwidth management platform 103 may use one or more data plane libraries and optimized network interface controller (NIC) drivers such as: a queue manager that implements lockless queues; a buffer manager that pre-allocates fixed size buffers; a memory manager that allocates pools of objects in memory and uses a ring to store free objects, which may ensure that objects are spread equally on all DRAM channels; poll mode drivers (PMD) that are designed to work without asynchronous notifications, which reduces overhead; and a packet framework (e.g., a set of libraries that help to develop packet processing).
  • NIC network interface controller
  • PMD poll mode drivers
  • bandwidth management platform 103 may use an abstraction layer that allows for implemented on different chips and systems.
  • bandwidth management platform 103 may include a set of functions, written according to the abstraction layer specification, that provides interfaces between an operating system and platform transport mechanism.
  • the set of functions may be generic such that bandwidth management platform 103 may be implemented on different chips and systems.
  • one or more of the components (and/or features/functions provided by those components) may be incorporated into the abstraction layer.
  • the compute devices lOla-lOln can be any of a variety of compute devices that can be operatively coupled to communication network 105.
  • a compute device lOla-lOln can be, for example, a personal computer, a tablet computer, a personal digital assistant (PDA), a cellular telephone, a Television, a portable/mobile internet device, a network enabled sensor (e.g., thermostat), a group of computing devices, servers, and/or any other device or group of devices capable of sending and/or receiving communication signals via a network.
  • the compute device lOla-lOln can include a web browser configured to access a webpage or website accessible via the gateway device 109 over communication network 105.
  • the compute device lOla-lOln can be configured to support, for example, HTML using JavaScript.
  • the compute device lOla-lOln can include a web browser, such as, Firefox, Safari, Dolphin, Opera and Chrome.
  • An Internet page or website (e.g., hosted by compute device 130) can be accessed by a user of a web browser at a compute device lOla-lOln by providing the web browser with a reference such as a uniform resource locator (URL), for example, of a webpage.
  • URL uniform resource locator
  • a user of a compute device lOla-lOln can access the compute device 130 via the gateway device 109 using a URL designated for the compute device 130.
  • compute device lOla-lOln can include specialized software for accessing a server (e.g., compute device 130) connected to the communication network 105 other than a browser, such as, for example, a specialized network-enabled application or program (e.g., FTP program, peer-to-peer file sharing software, online chat software, etc.).
  • a server e.g., compute device 130
  • a specialized network-enabled application or program e.g., FTP program, peer-to-peer file sharing software, online chat software, etc.
  • a compute device lOla-lOln can also include a display, monitor or user interface (not shown in FIG. 1), a keyboard, various ports (e.g., a USB port), and other user interface features, such as, for example, digital pens, mice, touch screen controls, audio components, and/or video components (each not shown).
  • a compute device lOla-lOln can be operatively coupled to communication network 105 via a user interface and network connections 111 and 113.
  • the memory 121a-121n of the compute device lOla-lOln can be, for example, a random access memory (RAM), a memory buffer, a hard drive, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), and/or so forth.
  • the memory 12 la- 12 In of the compute device lOla-lOln stores instructions to cause the processor 119a-119n to execute modules, processes, and/or functions associated with using a personal computer application, a mobile application, an internet web browser, a peer-to-peer network or file share, and/or the like.
  • the memory 121a-121n stores instructions to cause the processor 119a-l 19n to send signal information to a display (not shown in FIG. 1).
  • the processor 119a-119n of the compute device lOla-lOln can be any suitable processing device configured to run and/or execute a set of instructions or code.
  • the processor 119a-119n can be a general purpose processor, a central processing unit (CPU), an accelerated processing unit (APU), and/or the like.
  • the processor 119a-119n can be configured to run and/or execute a set of instructions or code stored in the memory 121a- 12 In associated with using a personal computer application, mobile application, an internet web browser, and/or the like.
  • the processor 119a-119n can run and/or execute a set of instructions associated with sending and/or receiving information to a server via the gateway device 109.
  • the compute device lOla-lOln can include a network interface (not shown in FIG. 1).
  • the network interface can be any suitable network interface configured to communicate with a LAN 107 and/or a communication network 105.
  • the network interface can be a network card configured to communicate with LAN 107 and/or the communication network 105 via a wired or wireless connection.
  • the gateway device 109 can function as, for example, a network management device.
  • the memory 125 included in the gateway device 109 can be, for example, a random access memory (RAM), a memory buffer, a hard drive, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), and/or so forth.
  • the memory 125 of the gateway device 109 includes a set of instructions executed by the processor 123, provided by the bandwidth management platform 103, to allocate a given amount of network bandwidth for use by the compute device 101a- 101 n.
  • the gateway device 109 can operatively couple the compute devices lOla-lOln associated with the LAN 107 to the communication network 105.
  • the processor 123 can be any suitable processor such as, for example, a general purpose processor, a central processing unit (CPU), a network processor, a front end processor, graphics processing unit (GPU), a fixed programmable gate array (FPGA), and/or the like. As such, the processor 123 is configured to perform a set of instructions stored in the memory 125.
  • FIG. 2 is a schematic illustration of a bandwidth management platform 200, according to an embodiment.
  • the bandwidth management platform 200 can be structurally and/or functionally similar to the bandwidth management platform 103 of FIG. 1. As shown in FIG.
  • the bandwidth management platform 200 can include a monitor module 201, a threshold calculation module 203, a use calculation module 205, a use identification module 207, a rate limiting module 209, and a data store 213. Furthermore, the bandwidth management platform 200 can communicate with other components of a computer network system (e.g., computer network system 100 of FIG. 1) via input signals 215 and output signals 217.
  • a computer network system e.g., computer network system 100 of FIG. 1
  • the bandwidth management platform 200 and its components may be located anywhere within a communication network system 100 including, but not limited to, within the gateway device 109, or in separate locations within the communication network system 100 of FIG. 1.
  • a module can be, for example, any assembly and/or set of operatively- coupled electrical components, and can include, for example, a memory, a processor, electrical traces, optical connectors, software (executing or to be executed in hardware) and/or the like. Furthermore, a module can be capable of performing one or more specific functions associated with the module, as discussed further below. In some embodiments, a module may be incorporated into an abstraction layer, cloud components, and/or an integrated circuit. For example, bandwidth management platform 200 may use an abstraction layer that allows for implemented on different chips and systems. For example, bandwidth management platform 200 may include a set of functions, written according to the abstraction layer specification, that provides interfaces between an operating system and platform transport mechanism.
  • bandwidth management platform 200 may be implemented on different chips and systems.
  • bandwidth management platform 200 may be implemented in cloud components (e.g., as described in FIG. 7 below).
  • bandwidth management platform 200 may access a set of functions/services available from cloud components.
  • bandwidth management platform 200 may be implemented in an integrated circuit.
  • the bandwidth management platform 200 can provide bandwidth management for compute devices (e.g., compute devices lOla-lOln of FIG. 1).
  • the bandwidth management platform 200 can be configured to communicate with the communication network 105 and can receive a signal from the compute devices lOla-lOln, from the gateway device 109, and/or any other device(s) in communication with the communication network 105 associated with a given network or communication protocol (e.g., Internet Protocol or a similar messaging or signaling system, Transmission Control Protocol (TCP), User Datagram Protocol (UDP), File Transfer Protocol (FTP), Simple Mail Transfer Protocol (SMTP), Hyper Text Transmission Protocol (HTTP), Internet Message Access Protocol (IMAP), Peer Distributed Transfer Protocol (PDTP), BitTorrent, and/or the like).
  • TCP Transmission Control Protocol
  • UDP User Datagram Protocol
  • FTP File Transfer Protocol
  • SMTP Simple Mail Transfer Protocol
  • HTTP Hyper Text Transmission Protocol
  • IMAP Internet Message Access Protocol
  • PDTP Peer Distributed Transfer Protocol
  • BitTorrent BitTorrent
  • the data store 213 can be, for example, a random access memory (RAM), a memory buffer, a hard drive, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a database and/or so forth.
  • the data store 213 can store data associated with bandwidth use of the compute devices lOla-lOln and/or data from the other portions of the bandwidth management platform 103, as described in further detail herein.
  • the monitor module 201 monitors network traffic between the LAN 107 and the communication network 105 (e.g., electronic signals associated with any of the aforementioned protocols) through, for example, an Internet gateway.
  • the monitor module 201 can periodically monitor a user’ s activities regarding bandwidth consumption.
  • the monitor module 201 can monitor network traffic (e.g., data packets) that passes through the gateway device 109 (e.g., between a compute device lOla-lOln and the communication network 105). The monitor module can then classify this data based on the source and/or destination compute device lOla-lOln of that data packet.
  • the monitor module 201 can determine bandwidth consumption of the link 117 (between the gateway device 109 and the communication network 105) on a per compute device lOla-lOln basis.
  • the monitor module 201 monitors and/or stores (e.g., in data store 213) a transient record to track metrics for a user of a compute device lOla-lOln. This record can be retained in the data store 213 for the duration of the session of the user and can be discarded when the user’s access and/or session is terminated.
  • the gateway device 109 can monitor and/or determine bandwidth consumption of the link 113 (between the gateway device 109 and the LAN 107) on a per compute device lOla-lOln basis.
  • the monitor module 201 can also monitor, collect and/or receive information associated with a total amount of bandwidth available between the compute devices lOla-lOln and the communication network 105. For example, the monitor module 201 can receive and/or compute information associated with a maximum amount of available bandwidth for a given time period (e.g., one hour). Similarly stated, the monitor module 201 can determine an amount of bandwidth the compute devices lOla-lOln can collectively use to access the communication network 105.
  • the monitor module 201 can store the monitored and/or collected data in the data store 213. For example, the monitor module 201 can send a signal to the data store 213 to store a maximum available bandwidth value. Additionally, the monitor module can send a signal to the data store 213 to store data associated with the bandwidth used by each compute device lOla-lOln, groups and/or categories of compute devices lOla-lOln, and/or the like. In other embodiments, the monitor module 201 can also provide this data to one or more of the other modules within the bandwidth management platform 200.
  • the threshold calculation module 203 can calculate one or more bandwidth usage thresholds that can be used to limit the bandwidth of specific compute devices lOla-lOln, as described in further detail herein. Specifically, the threshold calculation module 203 can receive the data associated with an available bandwidth for the compute devices lOla-lOln from the monitor module 201 and/or from the data store 213. Additionally, the threshold calculation module 203 can receive an indication of an amount of the available bandwidth being collectively used by the compute devices lOla-lOln for a specific time period from the monitor module 201 and/or from the data store 213.
  • the threshold calculation module 203 can also receive data associated with a number of compute devices lOla-lOln accessing the communication network 105 via the gateway device 109, a number of compute devices lOla-lOln classified in a specific usage category (as described in further detail herein), a number of compute devices lOla-lOln currently being rate limited, a percentage of bandwidth used by compute devices lOla-lOln classified in one or more usage categories, a priority associated with a compute device lOla-lOln (as described in further detail herein), and/or the like. [0058] Based on the received data, the threshold calculation module 203 can calculate one or more thresholds based on the received data.
  • the threshold calculation module 203 can calculate an upper threshold and/or a lower threshold.
  • the upper threshold can be calculated based on a percentage of total bandwidth used by the compute devices lOla-lOln classified in a fourth category (as described in further detail herein). In such a manner, the threshold calculation module 203 can calculate the upper threshold such that a percentage (e.g., 30%, 50%, 70%, etc.) of total bandwidth can be allocated to the compute devices lOla-lOln classified in the fourth category, as described in further detail herein.
  • the lower threshold can be calculated as a percentage of the upper threshold (e.g., 33%).
  • the threshold calculation module 203 can calculate the upper threshold and/or the lower threshold using data associated with a number of compute devices lOla-lOln accessing the communication network 105 via the gateway device 109, a number of compute devices lOla-lOln classified in a specific usage category (as further described in detail herein), a number of compute devices lOla-lOln currently being rate limited, a percentage of bandwidth used by compute devices lOla-lOln classified in one or more usage categories, a previous threshold, and/or the like.
  • the threshold calculation module 203 can store the threshold value(s) in the data store 213.
  • the thresholds can deviate variably depending on the number of compute devices lOla-lOln that are being throttled and also the total bandwidth capacity of the communication network 105 at a given time, as described in further detail herein. For example, a threshold can increase when a greater number of compute devices are rate limited. For another example, the threshold can decrease as the available network capacity increases.
  • the thresholds can be periodically and/or continually updated.
  • the threshold calculation module 203 can periodically and/or continuously receive data (as discussed above) and can revise and/or update the threshold values.
  • the threshold calculation module 203 can update and/or revise threshold values stored in data store 213.
  • the use calculation module 205 calculates information associated with bandwidth usage of the compute devices lOla-lOln. Specifically, the use calculation module 205 can receive data associated with bandwidth use of a compute device 101a- 10 In from the monitor module 201 and/or the data store 213.
  • the use calculation module 205 can calculate one or more usage values associated with the bandwidth use of the compute device lOla-lOln. For example, the use calculation module 205 can periodically and/or continuously calculate an average amount of bandwidth use for a compute device lOla-lOln, based on data provided by the monitor module 201.
  • the use calculation module 205 can periodically and/or continuously calculate an average amount of bandwidth use of a compute device (e.g., the compute device lOla-lOln) for a predetermined time period (e.g., 10 seconds).
  • a bandwidth use for a time period can periodically and/or continuously be determined by the use calculation module 205 for a compute device lOla-lOln for a second predetermined time period, less than the first predetermined time period (e.g., 1 second).
  • the use calculation module 205 can track bandwidth use for a specific compute device 101a- 101 n for several rolling windows of time such as, for example, a short term (10 seconds), a medium term (1 minute), a long term (5 minutes), etc. In some embodiments, the use calculation module 205 can store the usage values in the data store 213.
  • the use calculation module 205 can aggregate metrics collected for individual users of compute devices lOla-lOln by the monitor module 201. This enables the use calculation module 205 to determine the bandwidth load at a time and to track local maximums (high water mark) marking peak throughput on a periodic basis. Tracking the variability of local maximums by the use calculation module 205 enables the bandwidth management platform 200 to estimate actual maximum bandwidth service levels permitted by the gateway device 109.
  • the use calculation module 205 can store the calculation results in data store 213.
  • the use identification module 207 can assign a use category to each compute device lOla-lOln using the usage values associated with that compute device 101a- 10 In and the threshold values.
  • the use identification module 207 can receive the usage values for each compute device lOla-lOln from the use calculation module 205 and/or data store 213. Specifically, the use identification module 207 can receive usage values associated with one or more time periods (e.g., rolling windows) of bandwidth usage for each compute device lOla-lOln (or a subset of the compute devices lOla-lOln currently accessing the communication network 105).
  • time periods e.g., rolling windows
  • the use identification module 207 can receive one or more threshold values from the threshold calculation module 203. In some embodiments, the use identification module 207 can receive, for example, an upper threshold and a lower threshold. In other embodiments, the use identification module 207 can receive any number of threshold values. [0067] Using the usage value(s) and the threshold value(s), the use identification module 207 can classify the network traffic associated with (e.g., sent from and/or to) each compute device lOla-lOln based on the bandwidth usage characteristics as a function of total available bandwidth, for example, into multiple bandwidth use categories.
  • the use identification module 207 can group the compute devices lOla-lOln into multiple categories based on the usage values associated with each compute device lOla-lOln and the thresholds. In some embodiments, for example, the use identification module 207 can group the compute devices lOla-lOln into four categories. In some embodiments, the use identification module 207 can store an indication of a use value associated with a compute device lOla-lOln in the data store 213. [0068] In some embodiments, a first usage category can be defined to include compute devices lOla-lOln that use little or no bandwidth during a time period.
  • the first usage category is associated with a low activity or inactive usage pattern.
  • a compute device 101a- 101h is categorized in the first usage category when the compute device lOla-lOln is consuming little or no bandwidth over a period of time.
  • This activity type is typical of a compute device 101a- 101h at rest but with applications periodically polling a server (e.g., compute device 130) to check status, such as checking for new mail messages, or keeping telnet or Secure Shell (SSH) connections alive.
  • a server e.g., compute device 130
  • This usage pattern of the compute device lOla-lOln has little impact on the communication network 105, but can be occasionally sensitive to delay, for example, if receiving a response at the compute device lOla-lOln from an email server takes a long time.
  • the use identification module 207 classifies a compute device lOla-lOln in the first usage category if, for example, a usage value associated with bandwidth use for a short duration (e.g., 1 second) and a usage value associated with bandwidth use for a long duration (e.g., 1 minute) are both less than a lower threshold calculated by the threshold calculation module 203. Similarly stated, if the compute device lOla-lOln includes a first usage value and a second usage value less than the lower threshold, the use identification module 207 can identify a use pattern of the compute device lOla-lOln as being within the first usage category.
  • a second usage category can be defined to include compute devices lOla-lOln that use a large amount of bandwidth for a short period of time but use a smaller amount of bandwidth over a longer period of time.
  • the first bandwidth usage category can include electronic devices lOla-lOln that transmit and/or receive short, relatively infrequent bursts of data that use a large amount of bandwidth for a short period of time.
  • Such network traffic can be associated with, for example, surfing the World Wide Web, clicking on links to load pages, submitting forms, checking for new or unread e-mail, downloading e-mails with small to medium size attachments, and/or the like.
  • Such traffic uses a relatively high bandwidth usually for less than 10 seconds with a period of inactivity between bursts. This type of use is interactive, and round trip reply time is important for positive user experience. While the bandwidth used over a period of time is low, there are periodic bursts of very high bandwidth use lasting from, for example, one second to as much as, for example, 20 seconds.
  • the use identification module 207 classifies a compute device lOla-lOln in the second usage category if, for example, a usage value associated with bandwidth use for a short duration (e.g., 1 second) is greater than an upper threshold, but a usage value associated with bandwidth use for a long duration (e.g., 1 minute) is less than a lower threshold.
  • a usage value associated with bandwidth use for a short duration e.g., 1 second
  • a usage value associated with bandwidth use for a long duration e.g., 1 minute
  • the use identification module 207 can identify a use pattern of the compute device lOla-lOln as being within the second usage category.
  • a third usage category can be defined to include compute devices lOla-lOln that use a low to medium amount of bandwidth for both a short period of time and over a longer period of time.
  • the third usage category can include compute devices lOla-lOln that transmit and/or receive a relatively consistent low to medium amount of data over a long period of time. This can include network traffic associated with, for example, streaming audio and/or video, and/or using applications that have options to employ rate-controlled upload such as, for example, Dropbox.
  • Bandwidth use indicative of the third usage category is relatively constant for some period of time, but is self-limiting by its own nature.
  • the use identification module 207 classifies a compute device lOla-lOln in the third usage category if, for example, both a usage value associated with bandwidth use for a short duration (e.g., 1 second) and a usage value associated with bandwidth use for a longer duration (e.g., 1 minute) are greater than a lower threshold but less than an upper threshold.
  • the use identification module 207 can identify a use pattern of the compute device lOla-lOln as being within the third usage category.
  • a fourth usage category can be defined to include user devices lOla-lOln that use a high amount of bandwidth for both a short period of time and over a longer period of time.
  • Such network traffic can be associated with, for example, file transfers, peer-to- peer file sharing and/or downloading large mail attachments.
  • such network traffic can be associated with a large amount of data being transferred using a guaranteed protocol such as, for example, Transmission Control Protocol (TCP).
  • TCP Transmission Control Protocol
  • This type of traffic tends to use available bandwidth because of the nature of the TCP rate-matching algorithm. This type of traffic is not usually interactive nor is it real-time. Instead, a large amount of data is transferred, and the user has an expectation that data transfer will need time.
  • TCP The guaranteed delivery property of TCP causes the lost packets to be re-transmitted if they are lost or delayed. This type of bandwidth use may last from, for example, a minute to, for example, many hours until the transfer is completed. It takes only a few compute devices lOla-lOln using such applications to cause network congestion such that the experience of other users is negatively affected.
  • the use identification module 207 classifies a compute device lOla-lOln in the fourth usage category if, for example, both a usage value associated with bandwidth use for a short duration (e.g., 1 second) and a usage value associated with bandwidth use for a longer duration (e.g., 1 minute) are greater than a lower threshold and an upper threshold.
  • the use identification module 207 can identify a use pattern of the compute device lOla-lOln as being within the fourth usage category.
  • the use identification module 207 can classify a compute device lOla-lOln into any number of categories using any number of thresholds and/or usage values associated with that compute device lOla-lOln.
  • the rate limiting module 209 can limit and/or throttle data transfer through the gateway device 109 for one or more compute devices lOla-lOln based on the usage category of that compute device lOla-lOln. Specifically, the rate limiting module 209 can receive an indication of a use category assigned to a compute device lOla-lOln from the use identification module 207 and/or the data store 213. Based on the usage category, the rate limiting module 209 can limit and/or throttle data transfer for that compute device lOla-lOln.
  • the rate limiting module 209 can limit and/or throttle data transfer through the gateway device 109 for each compute device lOla-lOln classified in the fourth usage category. In such embodiments, the rate limiting module 209 can allow non rate limited data transfer through the gateway device 109 for the compute devices lOla-lOln classified in the first usage category, the second usage category or the third usage category.
  • the gateway device 109 can permit the network traffic of the first, second and third usage categories through the gateway without limiting the speed of the data transfer. For example, even though the bandwidth use of the compute devices lOla-lOln classified in the second usage category can be relatively high, the high bandwidth use occurs for a short period of time with a relatively long duration of time between bursts. Therefore, the overall impact on the network is minimal because the average bandwidth consumed over time is relatively low. For example, the stochastic nature of web surfing is such that the bandwidth of the type of traffic associated with the second usage category is not limited because requests for pages arc generally random and staggered across multiple users.
  • the rate limiting module 209 can limit and/or throttle data transfer through the gateway device 109 for compute devices lOla-lOln in categories other than the fourth category.
  • the rate limiting module 209 can limit and/or throttle data associated with compute devices lOla-lOln classified in the third category as well as the fourth category.
  • the rate limiting module 209 can limit and/or throttle data in the fourth category differently (e.g., at a different rate) than data in the third category.
  • the rate limiting module 209 can dynamically determine and/or calculate an amount to limit and/or throttle data transfer through the gateway device 109 for the compute devices lOla-lOln identified and/or selected to be rate limited. Specifically, the rate limiting module 209 can select a limited data transfer rate for compute devices lOla-lOln in specific categories based on any number of factors, such as, for example, the total capacity of the network, the number of compute devices lOla-lOln currently rate limited and/or priorities associated with those compute devices lOla-lOln (discussed further herein), the number of compute devices lOla-lOln in each priority classification that arc throttled, a ratio of compute devices lOla-lOln that are currently rate limited to the compute devices lOla-lOln that not currently rate limited, and/or the like.
  • factors such as, for example, the total capacity of the network, the number of compute devices lOla-lOln currently rate limited and
  • the data transfer rate for a compute device lOla-lOln that is being limited and/or throttled can be adjusted up or down as the total network traffic changes.
  • the data transfer rate and/or limit for compute devices lOla-lOln classified in the fourth usage category can increase as fewer compute devices lOla-lOln send data to the communication network 105 via the gateway device 109.
  • the rate limiting module 209 may not adjust the rate limit below a certain use (e.g., 32Kbps upload and 64Kbps download) even in the most dire network situations. If such low levels of bandwidth become chronic, it can be a sign that the communication network 105 is oversubscribed and bandwidth service levels need to be upgraded.
  • the rate limiting module 209 can select a data transfer rate for a compute device lOla-lOln that is being limited and/or throttled based on reserving a predetermined amount of bandwidth for compute devices lOla-lOln classified in a particular category.
  • the allocated bandwidth for compute devices lOla-lOln of the fourth category can be a predetermined percentage (e.g., between 30% and 70%) of the total available bandwidth.
  • the bandwidth for a particular compute device lOla-lOln in the fourth category can be determined based on the selected percentage and the total number of compute devices 101a- 101h in the fourth category.
  • the allocated bandwidth can be determined based, at least in part, on the number of compute devices lOla-lOln sending and/or receiving packets through the gateway device 109, such that as the number of users increases the rate limiting module 209 allocates a higher total bandwidth to the compute devices lOla-lOln classified in the fourth category.
  • the rate limiting module 209 can send to the threshold calculation module 203 an indication of a number of compute devices lOla-lOln classified in a specific usage category, a number of compute devices lOla-lOln currently being rate limited, a percentage of bandwidth used by compute devices lOla-lOln classified in one or more usage categories, and/or the like.
  • the threshold calculation module 203 can then recalculate the thresholds based on the updated data.
  • the use identification module 207 can classify the compute devices lOla-lOln differently as the overall conditions of the bandwidth usage through the gateway device 109 changes.
  • a compute device lOla-lOln can be assigned a priority value, which can affect the decision to rate limit that compute device lOla-lOln.
  • a priority value can be characterized with, for example, a number 1 to n. Higher priority classifications can be reserved for compute devices lOla-lOln, applications used, and/or users that should receive higher priority.
  • mission-critical systems and/or certain users can be assigned a high priority setting.
  • the priority setting can be used to determine the relative handling of compute devices lOla-lOln by the bandwidth management platform 200.
  • the priority assigned to a user or a compute device lOla-lOln can override the category in which the compute device lOla-lOln is classified. For example, a compute device lOla-lOln classified in the fourth category and with a high priority, might not be rate limited based on the high priority of that compute device lOla-lOln.
  • a high priority can decrease one or more threshold values that will be applied to the usage value for that compute device lOla-lOln.
  • the compute device lOla-lOln can use a greater amount of bandwidth prior to being classified in the fourth category.
  • the rate limiting module 209 can assign a greater percentage of the total bandwidth to use by higher priority compute devices lOla-lOln than for use you lower priority compute devices lOla-lOln.
  • network use data for users of compute devices lOla-lOln can be periodically measured by the monitor module 201, and users can be classified among different usage patterns by the use identification module 207 (as described in FIG. 6) based on the use data.
  • the decision to classify current use by a particular user of a compute device lOla-lOln can be dynamic and depend on network conditions (e.g., using the dynamic thresholds). Network conditions as a whole are subject to change from moment to moment, as well as individual usage patterns. Thus, the thresholds can change as network conditions change.
  • the use identification module 207 can decide how to classify any particular usage pattern depending on use relative to other use on the communication network 105 based on the dynamic thresholds.
  • a user of a compute device lOla-lOln may be classified differently depending on a combination of their own behavior combined with the network conditions as a whole, and a priority level assigned to the user and/or the compute device lOla-lOln.
  • two or more of the modules 201-209 of the bandwidth management platform 200 shown and described with respect to FIG. 2 can be combined.
  • the use identification module 207 and the rate limiting module 209 can be combined into a single module.
  • the indication of the use category assigned to a compute device lOla-lOln is not determined, sent and/or stored.
  • the combined module directly (e.g., without defining an indication of the use category) limits and/or throttles bandwidth based on comparing the usage values to the thresholds in a manner described above.
  • the use calculation module 205 can be included as part of the monitor module 201.
  • the system may use express path forwarding as opposed to TCP/IP (“Transmission Control Protocol Internet Protocol”) stack.
  • TCP/IP Transmission Control Protocol Internet Protocol
  • conventional systems may use a TCP/IP stack for communications.
  • the methods and systems may optimize packet flow (e.g., from an operating system to the Internet) by establishing an express path as opposed to managing and polling the different intercommunications (e.g., to determine if there is bandwidth).
  • the express path forwarding may comprise specialized rules and/or components to minimize the amount of copies made by the system. By doing so, the system minimizes the amount of copies needed to be manipulated.
  • the system may provide the express path forwarding through the use of several components including buffer management, pipelining of data, and unique combinations of code.
  • the system may include a software layer (e.g., which may be an abstraction layer) that is responsible for bringing data from one location (e.g., a physical disk, main memory, etc.) to another as needed.
  • the buffer manages the available memory and data flow by dividing the memory into a collection (e.g., a buffer pool).
  • the system may perform on-the-fly buffering. For example, due to the increased operating speed and efficiency the system may not need to divide memory up front, can now do on-the-fly buffering.
  • the system may also pipeline data by moving data from one component to another. The data may or may not be transformed, and it may be processed in real time (or streaming) instead of batches.
  • the system may streamline the packet flow by looking at a clock speed and cycle with a processor.
  • the system may also use IP filtering as opposed to IP tables.
  • IP tables may be used to manage network traffic.
  • the methods and systems may use dynamic filters to automatically move packets through the express path.
  • the system does not need to use IP tables because it may monitor an incoming and/or outcoming data stream itself using software (e.g., via deep packet inspection).
  • Deep packet inspection which is also known as DPI, information extraction, IX, or complete packet inspection, is a type of network packet filtering. Deep packet inspection evaluates the data part and the header of a packet that is transmitted through an inspection point, weeding out any non-compliance to protocol, spam, viruses, intrusions, and any other defined criteria to block the packet from passing through the inspection point.
  • the system may use deep packet inspection to determine if a particular packet is to be redirected to another destination, complies with data policing rules, etc.
  • the system may locate, detect, categorize, block, or reroute packets that have specific code or data payloads that are not detected, located, categorized, blocked, or redirected by conventional packet filtering. Unlike plain packet filtering, deep packet inspection goes beyond examining packet headers.
  • the system may evaluate the contents of a packet that is going through a checkpoint. Using rules that are assigned by the system, the system determines what to do with these packets in real time.
  • the system may use several techniques. For example, the system may use pattern or signature matching (e.g., the system may perform pattern or signature matching, analyzes each packet against a database of known network attacks).
  • the system may also use protocol anomaly (e.g., protocol definitions are used to determine which content should be allowed).
  • protocol anomaly e.g., protocol definitions are used to determine which content should be allowed.
  • the system may also use IPS and/or IDS solutions.
  • the system is able to check the contents of packets and then figure out where it came from, such as the service or application that sent it.
  • the system may monitor address and source ports in-line with IP filtering.
  • the system may apply filter rules to multiple lines or may apply different rules to each line.
  • the system may check each packet against each rule that is associated with a line. The rules are checked in a sequential process. After the system matches the packet to a rule, it stops the process and applies the matching rule.
  • the system may also use improved proxy.
  • conventional systems may use a proxy server to manage network traffic.
  • the methods and systems may add a proxy to the software, which may provide acceleration to a CPU through the operating system.
  • the system may terminate connections and issue a different type of congestion control.
  • the proxy protocol may be tuned to each application, the type of data, and/or the number of “ways” of communication (e.g., two-way audio (e.g., a voice call), two-way video (e.g., a video call)), one-way audio (e.g., radio), one-way video (e.g., online video streaming).
  • video may be more subject to issues with bandwidth, jitter, and/or synchronization, whereas audio is more susceptible to latency issues.
  • the system may select (e.g., from plurality of proxy protocols) and/or adjust a current proxy protocol (e.g., adjust a rule of the current proxy protocol) based on the “ways”, type, or application associated with the data transmission.
  • the system may also use a scheduler for dynamic adjustment of packet size. For example, while conventional systems may use a scheduler with multi-core systems, the methods and systems may implement a more efficient scheduler, which may dynamically adjust the packet size as the packets sizes are processed on any hardware chip, processor, etc.
  • the scheduler may schedule based on a packet size, amount of traffic, and/or type of traffic.
  • the scheduler may include logic which decides how (e.g., an in what order) to move packets.
  • the scheduler may be associated with a queuing system and/or storing packets temporarily until they are transmitted.
  • the system may have a single queue or multiple queues in which case each may hold the packets of one flow, classification, or priority.
  • the system may then schedule different processes on different cores.
  • the system may also allow for lower power consumption while providing enhanced processing capabilities such as throughput, packet forwarding, more web requests/second and/or the like.
  • the system may perform dynamic (on-the-fly) scheduling changes and adjust congestion control algorithms on the fly without disrupting the dataflow. The system is able to perform this feature as the data does not require sampling as described above.
  • FIGS. 3-5 are flowcharts of processes for providing dynamic bandwidth, according to various embodiments.
  • FIG. 3 shows a method 300 for allocating bandwidth based on a threshold value, according to an embodiment.
  • the threshold calculation module 203 receives (e.g., from monitor module 201) at a time T, a datum associated with use of a network (e.g., communication network 105) by a set of compute devices lOla-lOln using the communication network 105.
  • the datum may include, for example, average bandwidth use by each compute device 101 a- 101 n, a use profile identifier for a user, a bandwidth limit associated with a user of a compute device lOla-lOln, and/or the like.
  • the datum can also include a number of compute devices 101a- 10 In from the set of compute devices.
  • a monitor module 201 monitors use of the communication network 105 for each compute device lOla-lOln from the set of compute devices and can send the use data to the threshold calculation module 203.
  • method 300 may be performed by an abstraction layer, cloud components, and/or an integrated circuit.
  • method 300 may use an abstraction layer that allows for implemented on different chips and systems.
  • the abstraction layer may include a set of functions, written according to the abstraction layer specification, that provides interfaces between an operating system and platform transport mechanism.
  • the set of functions may be generic such that bandwidth management platform 200 may be implemented on different chips and systems.
  • method 300 may be implemented in cloud components (e.g., as described in FIG. 7 below).
  • method 300 may access a set of functions/services available from cloud components.
  • method 300 may be implemented in an integrated circuit.
  • the threshold calculation module 203 calculates a threshold value SI based, at least in part, on the datum and a bandwidth capacity C of the communication network 105 at the time T.
  • the threshold value SI can be based on data associated with a number of compute devices lOla-lOln accessing the communication network 105 via the gateway device 109, a number of compute devices lOla-lOln classified in a specific usage category (as described in further detail herein), a number of compute devices lOla-lOln currently being rate limited, a percentage of bandwidth used by compute devices lOla-lOln classified in one or more usage categories, a priority associated with a compute device lOla-lOln (as described in further detail herein), and/or the like.
  • the use identification module 207 receives, a usage value U associated with bandwidth use of a compute device lOla-lOln.
  • the use identification module 207 can receive the use value directly from each compute device lOla-lOln, and/or from the use calculation module 205.
  • the use value U can be associated with bandwidth use by the compute device lOla-lOln during a time period PI immediately preceding the time T.
  • the use identification module 207 can also receive the threshold value SI from the threshold calculation module 203.
  • a use calculation module 205 receives a signal (e.g., from a monitor module 201) associated with use of the communication network 105 by the compute device 101a- 101 n.
  • the use calculation module 205 can calculate the use value U based, at least in part, on the signal associated with use of the communication network 105 by the compute device lOla-lOln.
  • the use calculation module 205 can send the use value U to the use identification module 207 to be used for comparison, as described at step 307.
  • the use identification module 207 compares the use value U with the threshold value SI to determine a usage category associated with the compute device 101a- 101 n, or for a user of the compute device lOla-lOln.
  • a usage category can be defined based on type of network use by a compute device lOla-lOln.
  • a first category can identify network traffic associated with little or no bandwidth use during a time period
  • a second category can identify short, relatively infrequent bursts that use a large amount of bandwidth for a short period of time
  • a third category can identify network traffic associated with continuous use of relatively low bandwidth
  • a fourth category can identify network traffic associated with high bandwidth use for a relatively long duration.
  • the use categories can be predefined or added based on new user categories discovered from analyses of Big Data based on machine learning 702.
  • the rate limiting module 209 receives an indication of the usage category from the use identification module 207, for a compute device 101a- 101 n.
  • the rate limiting module 209 can limit bandwidth availability of the communication network 105 for the compute device 101a- 10 In based on the indication.
  • the system may use IP filtering and use a scheduler for dynamic adjustment of packet size. For example, if the compute device lOla-lOln is categorized in the second category (infrequent large amounts of bandwidth use), the rate limiting module 209 does not limit bandwidth use of the compute device lOla-lOln.
  • the rate limiting module 209 can limit bandwidth use of the compute device 101a- 101h (e.g., based on a rate calculated by the rate limiting module 209).
  • the rate limiting module 209 can limit bandwidth use of the communication network 105 at time T for a number of compute devices lOla-lOln from the set of compute devices included in the datum.
  • the rate limiting module 209 can limit bandwidth use of the communication network 105 for the compute device lOla-lOln based on various factors, in addition to the category indication. For example, the rate limiting module 209 can limit bandwidth based on a priority associated with the compute device lOla-lOln (for example, in a device profile or user profile). The rate limiting module 209 can also use other data to limit bandwidth.
  • the rate limiting module 209 can use number of compute devices lOla-lOln from the set of compute devices and for which the rate limiting module 209 is limiting bandwidth use of the communication network 105 at time T and/or number of compute devices lOla-lOln from the set of compute devices and for which the rate limiting module 209 is not limiting bandwidth use of the communication network 105 at time T.
  • the rate limiting module 209 can also use various statistical analysis results provided by the use calculation module 205 for limiting the bandwidth use such as, for example, a ratio of a number of compute devices lOla-lOln from the set of compute devices and for which the rate limiting module 209 is limiting bandwidth use of the communication network 105 at time T and a number of compute devices lOla-lOln from the set of compute devices and for which the rate limiting module 209 is not limiting bandwidth use of the communication network 105 at time T.
  • the threshold calculation module 203 can calculate a second threshold value S2 based, at least in part, on the datum and the bandwidth capacity of the communication network 105 at time T.
  • the use identification module 207 receives the first threshold value SI and the second threshold value S2 from the threshold calculation module 203.
  • the use identification module 207 also receives a use value associated with bandwidth use of the compute device lOla-lOln and associated with a second time period P2 immediately preceding time T (e.g., from the use calculation module 205).
  • the use identification module 207 can determine the usage category for the compute device lOla-lOln, or for a user of the compute device 101a- 101 n, based on the first threshold SI, the second threshold S2, the use value associated with the first time period PI , and the use value associated with the second time period P2.
  • FIG. 4 shows a method 400 for allocating bandwidth based on a usage category, according to one embodiment.
  • the system may use IP filtering and use a scheduler for dynamic adjustment of packet size.
  • the bandwidth management platform 200 receives, at a time Tl, a signal associated with use of a communication network 105 by a compute device lOla-lOln associated with a user.
  • the signal may include, for example, average bandwidth use by the compute device 101 a- 101 n, a use profile identifier for the user, a bandwidth limit associated with the user, etc.
  • method 400 may be performed by an abstraction layer, cloud components, and/or an integrated circuit.
  • method 300 may use an abstraction layer that allows for implemented on different chips and systems.
  • the abstraction layer may include a set of functions, written according to the abstraction layer specification, that provides interfaces between an operating system and platform transport mechanism.
  • the set of functions may be generic such that bandwidth management platform (e.g., bandwidth management 200 (FIG. 2)) may be implemented on different chips and systems.
  • bandwidth management platform e.g., bandwidth management 200 (FIG. 2)
  • method 400 may be implemented in cloud components (e.g., as described in FIG. 7 below).
  • method 400 may access a set of functions/services available from cloud components.
  • method 400 may be implemented in an integrated circuit.
  • the use calculation module 205 calculates a first usage value U1 based, at least in part, on the signal associated with use of the communication network 105.
  • the first usage value U1 is calculated based on bandwidth use of the user for a first time period PI immediately preceding time Tl.
  • the use calculation module 205 calculates a second usage value U2 based, at least in part, on the signal associated with use of the communication network 105.
  • the second usage value U2 is calculated based on bandwidth use of the user for a second time period P2 immediately preceding time Tl.
  • the first time period PI is different from the second time period P2.
  • the use identification module 207 compares the first usage value U1 and the second usage value U2 with a value of at least one threshold Y1 associated with use of the communication network 105 at time Tl (the threshold value Y1 can be, for example, provided by the threshold calculation module 203).
  • the use identification module 207 can determine a usage category associated with the user based on the comparison, as described above with regards to step 307 of FIG. 3.
  • the value of threshold Y1 can be calculated based, at least in part, on a bandwidth capacity of the communication network 105.
  • the value of threshold Y1 can also be calculated based on a number of users using the communication network 105 and/or priorities associated with those users.
  • the threshold value calculated by the threshold calculation module 203 can be different at different times. For example, a value of the at least one threshold Y2 at a second time T2 after time Tl can be different from the value of the at least one threshold Y1 at time Tl . Because the bandwidth capacity of the communication network 105, the number of users using the communication network 105, the users’ priority and the users’ category can dynamically change, a value of a threshold that is calculated based on the bandwidth capacity of the communication network 105, the number of users using the communication network 105, the users’ priority and the users’ category, can also dynamically change.
  • the threshold calculation module 203 calculates the value of the at least one threshold Y1 (or Y2) based, at least in part, on a number of users using the communication network 105, for which bandwidth use of the communication network 105 is being limited at time T1 (or T2) and/or a number of users using the communication network 105 for which bandwidth use of the communication network 105 is not being limited at time T1 (or T2).
  • the threshold calculation module 203 can calculate a threshold value based on a ratio of a number of users using the communication network 105, for which bandwidth use of the communication network 105 is being limited at time T1 (or T2) and a number of users using the communication network 105 for which bandwidth use of the communication network 105 is not being limited at time T1 (or T2).
  • the bandwidth management platform 200 sends a signal (e.g., to a rate limiting module 209) to limit bandwidth use of the user (e.g., a compute device lOla-lOln associated with the user) of the communication network 105 for a time period P3 after time T based, at least in part, on the usage category and the threshold Y 1 (and/or Y2).
  • the signal to limit bandwidth use can be additionally based on a priority associated with the user or associated with the compute device lOla-lOln (for example, identified in a device profile or user profile).
  • FIG. 5 shows a method 500 for allocating bandwidth based on a threshold and a use value, according to an embodiment.
  • the system may use IP filtering and use a scheduler for dynamic adjustment of packet size.
  • the bandwidth management platform 200 receives, at a first time, a first datum associated with use of a communication network 105 by a set of compute devices lOla-lOln.
  • the first datum may include, for example, average bandwidth use by each compute device 101 a- 101 n, a use profile identifier for a user of a compute device 101a- 10 In, a bandwidth limit associated with a user of a compute device lOla-lOln, etc.
  • the first datum can also include a number of compute devices lOla-lOln from the set of compute devices.
  • method 500 may be performed by an abstraction layer, cloud components, and/or an integrated circuit.
  • method 300 may use an abstraction layer that allows for implemented on different chips and systems.
  • the abstraction layer may include a set of functions, written according to the abstraction layer specification, that provides interfaces between an operating system and platform transport mechanism.
  • the set of functions may be generic such that bandwidth management platform (e.g., bandwidth management 200 (FIG. 2)) may be implemented on different chips and systems.
  • FOG. 2 bandwidth management 200
  • method 500 may be implemented in cloud components (e.g., as described in FIG. 7 below).
  • method 500 may access a set of functions/services available from cloud components.
  • method 500 may be implemented in an integrated circuit.
  • the threshold calculation module 203 calculates a first threshold value based, at least in part, on the first datum and a bandwidth capacity of the communication network 105 at the first time.
  • the first datum can include a number of compute devices lOla-lOln from the set of compute devices using the communication network 105 and/or any other data associated with bandwidth use of the network, as described above.
  • the bandwidth management platform 200 receives a first usage value associated with bandwidth use of the communication network 105 of a compute device lOla-lOln from the set of compute devices for a time period immediately preceding the first time.
  • the bandwidth management platform 200 can determine a limitation for bandwidth use based on the data received and the values received and/or calculated at steps 501-505.
  • the bandwidth management platform 200 sends, at a second time after the first time a signal to limit bandwidth use in the communication network 105 of the compute device lOla-lOln based, at least in part, on the first threshold and the first usage value.
  • the signal to limit bandwidth use can limit bandwidth use based, at least in part, on a priority associated with the compute device lOla-lOln.
  • the first datum can include a number of compute devices lOla-lOln from the set of compute devices for which the signal to limit bandwidth use is limiting bandwidth use of the communication network 105 at the first time.
  • the bandwidth management platform 200 receives, at a third time after the second time a second datum associated with use of the communication network 105 by the set of compute devices using the communication network 105.
  • the threshold calculation module 203 calculates a second threshold value based, at least in part, on the second datum and a bandwidth capacity of the communication network 105 at the third time. The second threshold value is different from the first threshold value.
  • the bandwidth management platform 200 receives a second usage value associated with bandwidth use of the communication network 105 of the compute device 101a- 10 In from the set of compute devices for a time period immediately preceding the second time.
  • the time period immediately preceding the first time can have a duration substantially equal to the duration of the time period immediately preceding the second time.
  • the time period immediately preceding the first time can be different from the time period immediately preceding the second time.
  • the second usage value can be substantially equal to the first usage value.
  • the bandwidth management platform 200 can determine not to limit bandwidth use for a compute device lOla-lOln.
  • the bandwidth management platform 200 sends, at a fourth time after the third time a signal to not limit bandwidth use in the communication network 105 of the compute device lOla-lOln based, at least in part, on the second threshold and the second usage value.
  • the components 201-209 of the bandwidth management platform 200 can store received data, calculated values, etc. in the data store 213.
  • the decision to rate limit the compute device lOla-lOln at the second time but not at the fourth time is based on the dynamic threshold changing based on the characteristics of the network.
  • the threshold calculation module 203 calculate the first threshold value based, at least in part, on a number of compute devices lOla-lOln from the set of compute devices for which bandwidth use of the communication network 105 is being limited at the first time, and/or a number of compute devices lOla-lOln from the set of compute devices for which bandwidth use of the communication network 105 is not being limited at the first time.
  • the threshold calculation module 203 can calculate the first threshold based, at least in part, on a ratio of (1) a number of compute devices lOla-lOln from the set of compute devices for which bandwidth use of the communication network 105 is being limited at the first time, and (2) a number of compute devices lOla-lOln from the set of compute devices for which bandwidth use of the communication network 105 is not being limited at the first time.
  • the methods 300, 400, 500 shown and described in FIGS. 3-5 are shown as having a “start” and an -end", in some embodiments, the methods 300, 400, 500 illustrate a one or more iterations of a periodic and/or continuous process.
  • the method 300 can return to step 301 and perform the method 300 to further assess the compute device and its usage category at a subsequent time.
  • method 500 of FIG. 5 illustrates two iterations of a process. In the first iteration and at step 507, the bandwidth of the compute device is limited. In the second iteration and at step 515, the bandwidth of the compute device is not limited. Such iterations can continuously and/or periodically reassess the bandwidth use of a compute device, the total amount of bandwidth being used by the network, the amount of bandwidth available, and/or the like.
  • FIG. 6 is an exemplary diagram of dynamic bandwidth allocation, according to an embodiment.
  • the graph of FIG. 6 shows a number of various usage patterns as previously described.
  • the uses 601 and 603 are associated by compute devices that are being throttled dynamically as they exhibit prolonged high-bandwidth use behavior.
  • use 601 can be related to initializing a video stream that is being limited affecting download speed but not the actual viewing of the stream.
  • the usage pattern 603 can be related to a movie download.
  • limiting use 603 can affect the speed of the download.
  • the system may use a proxy protocol that is tuned to each application, the type of data, and/or the number of “ways” of communication (e.g., two-way audio (e.g., a voice call), two- way video (e.g., a video call)), one-way audio (e.g., radio), one-way video (e.g., online video streaming).
  • two-way audio e.g., a voice call
  • two- way video e.g., a video call
  • one-way audio e.g., radio
  • video e.g., online video streaming.
  • video may be more subject to issues with bandwidth, jitter, and/or synchronization, whereas audio is more susceptible to latency issues.
  • Limiting uses 601 and 603, allows the bandwidth management platform 200 to allocate higher levels of bandwidth to compute devices with uses 605 and 607.
  • the compute devices associated with uses 605 and 607 can experience a higher level of performance than they would have had if the uses 601 and 603 were not limited.
  • Use 605, for example, can represent video calls (e.g., Skype calls) and basic network monitoring tools and will not be affected by the bandwidth limitations.
  • Use 607 can represent bursting and/or streaming traffic that can remain unaffected by bandwidth limitations because bulk traffic has been limited.
  • Embodiments described herein discuss a bandwidth management platform 103 that can limit bandwidth between a compute device lOla-lOln and a communication network 105.
  • the bandwidth management platform 103 can limit data traffic sent from a compute device lOla-lOln to communication network 105 (upstream traffic) and/or data traffic received from communication network and destined to the compute device lOla-lOln (downstream traffic).
  • upstream traffic can be independently monitored and controlled from the downstream traffic.
  • the monitor module 201 can separately monitor the upstream traffic and the downstream traffic associated with a compute device 10 la-10 In.
  • the threshold calculation module 203 FIG.
  • the rate limiting module 209 can rate limit the downstream traffic of a compute device lOla-lOln without rate limiting the upstream traffic of that compute device lOla-lOln.
  • bandwidth management platform 103 of FIG. 1 is shown and described as being between a LAN 107 and a communication network 105, in other embodiments the bandwidth management platform can be communicatively positioned between any two devices or groups of devices. Similarly stated, the bandwidth management platform can be placed in the communication path between any two devices or groups of devices. For example, in some embodiments, the bandwidth management platform can be positioned between two disparate and separate communication networks. For another example, the bandwidth management platform can be embedded in a single communication network that has been configured such that the data traffic between a specific compute device passes through the bandwidth management platform.
  • FIG. 7 is an exemplary diagram of a system for a dynamic bandwidth allocation platform, according to an embodiment.
  • system 700 may represent the components used for implementing dynamic bandwidth allocation platform.
  • system 700 may include mobile device 722 and user terminal 724. While shown as a smartphone and personal computer, respectively, in FIG. 7, it should be noted that mobile device 722 and user terminal 724 may be any computing device, including, but not limited to, a laptop computer, a tablet computer, a hand-held computer, other computer equipment (e.g., a server), including “smart,” wireless, wearable, and/or mobile devices.
  • FIG. 7 also includes cloud components 710.
  • Cloud components 710 may alternatively be any computing device as described above and may include any type of mobile terminal, fixed terminal, or other device.
  • cloud components 710 may be implemented as a cloud computing system and may feature one or more component devices.
  • system 700 is not limited to three devices. Users may, for instance, utilize one or more other devices to interact with one another, one or more servers, or other components of system 700.
  • one or more operations are described herein as being performed by particular components of system 700, those operations may, in some embodiments, be performed by other components of system 700.
  • components of mobile device 722 those operations may, in some embodiments, be performed by components of cloud components 710.
  • the various computers and systems described herein may include one or more computing devices that are programmed to perform the described functions. Additionally or alternatively, multiple users may interact with system 700 and/or one or more components of system 700.
  • each of these devices may receive content and data via input/output (hereinafter “I/O”) paths.
  • I/O input/output
  • Each of these devices may also include processors and/or control circuitry to send and receive commands, requests, and other suitable data using the EO paths.
  • the control circuitry may comprise any suitable processing circuitry.
  • Each of these devices may also include a user input interface and/or user output interface (e.g., a display) for use in receiving and displaying data.
  • a user input interface and/or user output interface e.g., a display
  • both mobile device 722 and user terminal 724 include a display upon which to display data.
  • mobile device 722 and user terminal 724 are shown as touchscreen smartphones, these displays also act as user input interfaces.
  • the devices may have neither user input interface nor displays and may instead receive and display content using another device (e.g., a dedicated display device such as a computer screen and/or a dedicated input device such as a remote control, mouse, voice input, etc.).
  • the devices in system 700 may run an application (or another suitable program). The application may cause the processors and/or control circuitry to perform operations related to dynamic bandwidth allocation platform.
  • Each of these devices may also include electronic storages.
  • the electronic storages may include non-transitory storage media that electronically stores information.
  • the electronic storage media of the electronic storages may include one or both of (i) system storage that is provided integrally (e.g., substantially non-removable) with servers or client devices or (ii) removable storage that is removably connectable to the servers or client devices via, for example, a port (e.g., a USB port, a firewire port, etc.) or a drive (e.g., a disk drive, etc.).
  • a port e.g., a USB port, a firewire port, etc.
  • a drive e.g., a disk drive, etc.
  • the electronic storages may include one or more of optically readable storage media (e.g., optical disks, etc.), magnetically readable storage media (e.g., magnetic tape, magnetic hard drive, floppy drive, etc.), electrical charge-based storage media (e.g., EEPROM, RAM, etc.), solid-state storage media (e.g., flash drive, etc.), and/or other electronically readable storage media.
  • the electronic storages may include one or more virtual storage resources (e.g., cloud storage, a virtual private network, and/or other virtual storage resources).
  • the electronic storages may store software algorithms, information determined by the processors, information obtained from servers, information obtained from client devices, or other information that enables the functionality as described herein.
  • FIG. 7 also includes communication paths 728, 730, and 732.
  • Communication paths 728, 730, and 732 may include the Internet, a mobile phone network, a mobile voice or data network (e.g., a 5G or LTE network), a cable network, a public switched telephone network, or other types of communications network or combinations of communications networks.
  • Communication paths 728, 730, and 732 may separately or together include one or more communications paths, such as a satellite path, a fiber-optic path, a cable path, a path that supports Internet communications (e.g., IPTV), free-space connections (e.g., for broadcast or other wireless signals), or any other suitable wired or wireless communications path or combination of such paths.
  • the computing devices may include additional communication paths linking a plurality of hardware, software, and/or firmware components operating together.
  • the computing devices may be implemented by a cloud of computing platforms operating together as the computing devices.
  • Cloud components 710 include machine learning model 702.
  • Machine learning model 702 may take inputs 704 and provide outputs 706.
  • the inputs may include multiple data sets such as a training data set and a test data set.
  • cloud components 710 may use the efficiencies and higher granularity of data observation to analyze, systematically extract information from, and/or otherwise deal with data sets that are too large or complex to be dealt with by traditional data-processing application software (e.g., Big Data).
  • Cloud components 710 may also characterize actual data (i.e., not sampled data) to train models (e.g., machine learning model 702) and perform analyses on larger data sets and/or in near real-time or real-time.
  • Each of the plurality of data sets e.g., inputs 704 may include data subsets related to user data, contact strategies, and results.
  • outputs 706 may be fed back to machine learning model 702 as input to train machine learning model 702 (e.g., alone or in conjunction with user indications of the accuracy of outputs 706, labels associated with the inputs, or with other reference feedback information).
  • machine learning model 702 may update its configurations (e.g., weights, biases, or other parameters) based on the assessment of its prediction (e.g., outputs 706) and reference feedback information (e.g., user indication of accuracy, reference labels, or other information).
  • connection weights may be adjusted to reconcile differences between the neural network’s prediction and the reference feedback.
  • one or more neurons (or nodes) of the neural network may require that their respective errors are sent backward through the neural network to facilitate the update process (e.g., backpropagation of error).
  • Updates to the connection weights may, for example, be reflective of the magnitude of error propagated backward after a forward pass has been completed. In this way, for example, the machine learning model 702 may be trained to generate better predictions.
  • machine learning model 702 may include an artificial neural network.
  • machine learning model 702 may include an input layer and one or more hidden layers.
  • Each neural unit of machine learning model 702 may be connected with many other neural units of machine learning model 702. Such connections can be enforcing or inhibitory in their effect on the activation state of connected neural units.
  • each individual neural unit may have a summation function which combines the values of all of its inputs together.
  • each connection (or the neural unit itself) may have a threshold function such that the signal must surpass before it propagates to other neural units.
  • Machine learning model 702 may be self-learning and trained, rather than explicitly programmed, and can perform significantly better in certain areas of problem solving, as compared to traditional computer programs.
  • an output layer of machine learning model 702 may correspond to a classification of machine learning model 702 and an input known to correspond to that classification may be input into an input layer of machine learning model 702 during training.
  • an input without a known classification may be input into the input layer, and a determined classification may be output.
  • machine learning model 702 may include multiple layers (e.g., where a signal path traverses from front layers to back layers). In some embodiments, back propagation techniques may be utilized by machine learning model 702 where forward stimulation is used to reset weights on the “front” neural units. In some embodiments, stimulation and inhibition for machine learning model 702 may be more free-flowing, with connections interacting in a more chaotic and complex fashion. During testing, an output layer of machine learning model 702 may indicate whether or not a given input corresponds to a classification of machine learning model 702 (e.g., whether or not network communications correspond to a rule, whether or not to enforce a rule, etc.).
  • a classification of machine learning model 702 e.g., whether or not network communications correspond to a rule, whether or not to enforce a rule, etc.
  • Hardware modules can include, for example, a general-purpose processor, a field programmable gate array (FPGA), and/or an application specific integrated circuit (ASIC).
  • Software modules (executed on hardware) can be expressed in a variety of software languages (e.g., computer code), including C, C++, JavaTM, Ruby, Visual BasicTM, and other object-oriented, procedural, or other programming language and development tools.
  • Examples of computer code include, but are not limited to, micro code or micro-instructions, machine instructions, such as, for example, produced by a compiler, code used to produce a web service, and files containing higher-level instructions that are executed by a computer using an interpreter. Additional examples of computer code include, but are not limited to, control signals, encrypted code, and compressed code.
  • Some embodiments described herein relate to a computer storage product with a non- transitory computer-readable medium (also can be referred to as a non-transitory processor- readable medium) having instructions or computer code thereon for performing various computer- implemented operations.
  • the computer-readable medium (or processor-readable medium) is non- transitory in the sense that it does not include transitory propagating signals per se (e.g., a propagating electromagnetic wave carrying information on a transmission medium such as, for example, space or a cable).
  • the media and computer code also can be referred to as code
  • code can be those designed and constructed for the specific purpose or purposes.
  • non-transitory computer-readable media include, but are not limited to, magnetic storage media such as, for example, hard disks, floppy disks, and magnetic tape; optical storage media such as, for example, Compact Disc/Digital Video Discs (CD/DVDs), Compact Disc-Read Only Memories (CD- ROMs), and holographic devices; magneto-optical storage media such as, for example, optical disks; carrier wave signal processing modules; and hardware devices that are specially configured to store and execute program code, such as, for example, Application-Specific Integrated Circuits (ASICs), Programmable Logic Devices (PLDs), Read-Only Memory (ROM) and Random-Access Memory (RAM) devices.
  • ASICs Application-Specific Integrated Circuits
  • PLDs Programmable Logic Devices
  • ROM Read-Only Memory
  • RAM Random-Access Memory

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

Dans certains modes de réalisation, une quantité de consommation de bande passante agrégée pour un ensemble de dispositifs informatiques sur un réseau peut être déterminée sur une première période de temps. Une quantité de bande passante disponible sur le réseau peut être déterminée sur la première période de temps. Un rapport de la quantité de consommation de bande passante agrégée à la quantité de bande passante disponible peut être déterminé. Un seuil de bande passante peut être déterminé sur la base du rapport. Une limite de débit pour un premier dispositif informatique (de l'ensemble de dispositifs informatiques) peut être définie sur la base d'une comparaison du seuil de bande passante à la consommation de bande passante du premier dispositif informatique.
PCT/US2021/061176 2020-11-30 2021-11-30 Systèmes et procédés pour une attribution et une optimisation dynamiques de bande passante WO2022115774A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US202063119366P 2020-11-30 2020-11-30
US63/119,366 2020-11-30

Publications (1)

Publication Number Publication Date
WO2022115774A1 true WO2022115774A1 (fr) 2022-06-02

Family

ID=81753688

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US2021/061176 WO2022115774A1 (fr) 2020-11-30 2021-11-30 Systèmes et procédés pour une attribution et une optimisation dynamiques de bande passante

Country Status (1)

Country Link
WO (1) WO2022115774A1 (fr)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050243862A1 (en) * 1996-07-02 2005-11-03 Microsoft Corporation Adaptive bandwidth throttling for network services
US20070016688A1 (en) * 2005-07-15 2007-01-18 Microsoft Corporation Background network bandwidth sharing behind gateway devices
US20130107716A1 (en) * 2011-10-28 2013-05-02 Calxeda, Inc. System and method for dynamically power and performance optimized server interconnects
US20150029864A1 (en) * 2013-07-25 2015-01-29 Grigore Raileanu System and method for managing bandwidth usage rates in a packet-switched network
US20200267070A1 (en) * 2012-04-13 2020-08-20 CirrusWorks, Inc. Systems and methods for setting a rate limit for a computing device

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050243862A1 (en) * 1996-07-02 2005-11-03 Microsoft Corporation Adaptive bandwidth throttling for network services
US20070016688A1 (en) * 2005-07-15 2007-01-18 Microsoft Corporation Background network bandwidth sharing behind gateway devices
US20130107716A1 (en) * 2011-10-28 2013-05-02 Calxeda, Inc. System and method for dynamically power and performance optimized server interconnects
US20200267070A1 (en) * 2012-04-13 2020-08-20 CirrusWorks, Inc. Systems and methods for setting a rate limit for a computing device
US20150029864A1 (en) * 2013-07-25 2015-01-29 Grigore Raileanu System and method for managing bandwidth usage rates in a packet-switched network

Similar Documents

Publication Publication Date Title
US11196653B2 (en) Systems and methods for dynamic bandwidth allocation and optimization
US11582163B2 (en) System for early system resource constraint detection and recovery
US10262700B2 (en) System and method for determining a cause of network congestion
US10587583B2 (en) Prioritizing application traffic through network tunnels
US9270600B2 (en) Low-latency lossless switch fabric for use in a data center
Hwang et al. Deadline and incast aware TCP for cloud data center networks
Xu et al. Fairness in fog networks: Achieving fair throughput performance in MQTT-based IoTs
Ali et al. Improved End-to-end service assurance and mathematical modeling of message queuing telemetry transport protocol based massively deployed fully functional devices in smart cities
US10862820B2 (en) Method for prioritizing network packets at high bandwidth speeds
Iqbal et al. Instant queue occupancy used for automatic traffic scheduling in data center networks
Patel et al. Comparative analysis of congestion control algorithms using ns-2
George et al. Congestion control mechanism for unresponsive flows in internet through active queue management system (AQM)
WO2022115774A1 (fr) Systèmes et procédés pour une attribution et une optimisation dynamiques de bande passante
Salah Modeling and analysis of PC-based software routers
Ma et al. Intelligent queue management of open vSwitch in multi-tenant data center
Lau et al. Optimizing the performance of OpenFlow Protocol over QUIC
KR20200108293A (ko) 네트워크 중립성 시험을 위한 시스템 및 방법
Gao et al. Hard Limit CoDel: An Extension of the CoDel Algorithm
Deng et al. BBR-based and fairness-guaranteed congestion control and packet scheduling for MPQUIC over heterogeneous networks
Wang et al. An Incast-Coflow-Aware Minimum-Rate-Guaranteed Congestion Control Protocol for Datacenter Applications
Dai et al. SP-DG: A programmable packet-level scheduling for queuing delay guarantees in time-critical networks
Grosvenor Latency-First datacenter network scheduling
CN118264620A (en) Computing system and method for flexible load balancing over multipath networks
Nakashima Queue management for the heavy-tailed traffics
Kansanen Wide area network acceleration in corporate networks

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21899220

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21899220

Country of ref document: EP

Kind code of ref document: A1