US20070263616A1 - Increasing link capacity via traffic distribution over multiple WI-FI access points - Google Patents

Increasing link capacity via traffic distribution over multiple WI-FI access points Download PDF

Info

Publication number
US20070263616A1
US20070263616A1 US11433359 US43335906A US2007263616A1 US 20070263616 A1 US20070263616 A1 US 20070263616A1 US 11433359 US11433359 US 11433359 US 43335906 A US43335906 A US 43335906A US 2007263616 A1 US2007263616 A1 US 2007263616A1
Authority
US
Grant status
Application
Patent type
Prior art keywords
access points
device
directing
available
method
Prior art date
Legal status (The legal status 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 status listed.)
Abandoned
Application number
US11433359
Inventor
Paul Castro
Frederique Giraud
Danny Yeh
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
International Business Machines Corp
Original Assignee
International Business Machines Corp
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

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L12/00Data switching networks
    • H04L12/28Data switching networks characterised by path configuration, e.g. local area networks [LAN], wide area networks [WAN]
    • H04L12/2854Wide area networks, e.g. public data networks
    • H04L12/2856Access arrangements, e.g. Internet access
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/70Admission control or resource allocation
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/70Admission control or resource allocation
    • H04L47/78Resource allocation architecture
    • H04L47/788Autonomous allocation of resources
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/70Admission control or resource allocation
    • H04L47/80Actions related to the nature of the flow or the user
    • H04L47/803Application aware
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04LTRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00Traffic regulation in packet switching networks
    • H04L47/70Admission control or resource allocation
    • H04L47/82Miscellaneous aspects
    • H04L47/821Prioritising resource allocation or reservation requests

Abstract

A method and system for combining the bandwidth available to a user communications device from multiple access points by switching output requests and corresponding replies to such requests among a plurality of such access points. In the simplest implementation the switching can be on a round-robin basis, but the switching can be enhanced to prioritize and use quality of service requirements associated with each output request, and also to measure congestion at each access point and base switching upon the congestion level. The invention operates by generating subset ISO communication stacks for each available access point, preferably the subset from the IP layer to the data control link, and switching between these stacks.

Description

    BACKGROUND OF THE INVENTION
  • 1. Field of the Invention
  • The present invention generally relates to distribution of communications traffic and, more particularly, to efficient use of multiple network access points available to a communications device.
  • 2. Background Description
  • In a client connected to the Internet, the communication bottleneck is usually the link from the computer to the access point. Whether the link is a wired Ethernet card or Wi-Fi (Wireless Ethernet) compatible link, the access point hub is likely shared by many users, so that the effective bandwidth used by the client is a small fraction of the traffic passing through the hub.
  • Furthermore, in most operating systems one and only one link is used by all the Internet applications running on the client. When downloading data from the Internet, especially when graphic or video files are involved, this reliance upon a single link leads to long queues at the access point, thus worsening the resulting delays.
  • Let us consider the current situation with Windows and Unix operating systems. In general, a laptop has both an Ethernet adapter and a wireless adapter to connect to the Internet. But only one access point is selected and used to the exclusion of the other, regardless of the actual total bandwidth available. Furthermore in the Wi-Fi case, a single access point is automatically selected among a possibly large number of available access points. But in the case of the wired Ethernet adapter, you can only connect to one access point, whereas in the wireless case you can choose the access point. And there is no dynamic allocation of physical links to applications.
  • Take the example of the browser. It submits requests and receives replies through a single access point statically configured. The same is true for electronic mail, FTP and all downloading applications in general. They talk to the Internet through a single specified standard port assigned during configuration of the application. This port in turn is assigned by the OS to the single physical link.
  • Most of the known solutions to this problem are hardware based. Some of the proposed solutions involve the design and implementation of proprietary schemes. There are a number of drawbacks to these known and proposed solutions. Most of the solutions involve the modification to the physical link standards, often adding functionality. Some require the turning on and off of link adapters which are not used simultaneously. This creates delays and only one access point is available at a time. This is advantageous only when the link capacity offered by each access point is large compared to the data traffic passing through the link. This is not the case when the link capacity is the bottleneck of the data traffic passing through such link.
  • Examples of prior art approaches include the disclosures of U.S. Patent Publication Number US20040223469A1to Bahl et al., “System and method for concurrent operation of a wireless device in two disjoint wireless networks”. This is hardware based and does switching from one wireless network to the other. U.S. Patent Publication Number US20050165916A1 to Cromer et al., “System and method for concurrent WLAN and WPAN wireless modes from a single device,” describes a way to switch between a Wi-Fi wireless network and a wireless cellular phone/pda network. U.S. Patent Publication Number US20040218580A1 to Bahl et al., “Method to enable simultaneous connections to multiple wireless network using a single radio,” describes a way of turning on and off adapters to different wireless networks, but only one network is activated at any given time.
  • There are a number of prior art references that address routing, traffic congestion, and traffic monitoring technologies. U.S. Patent Publication Number US20030002444 to Shin et al., “Route determining method in a multi protocol label switching (MPLS) network,” describes an autonomous network where traffic is sped up through routers by specifying a path for a sequence of packets so routers do not need to look up the address of the next node, as for example in an IP Virtual Private Network. This patent application describes a scheme where the input label edge router of such network calculates one or more desirable routes from the source to the destination given the interlink capacity between the nodes of the given network. The input edge router then partitions and allocates the incoming traffic between these calculated routes. While this scheme may look attractive in theory, it would not work well in practice because the calculations are based on transient data due to the probabilistic nature of the link capacity between nodes. Further, clients would have to constantly receive data from all the relevant links, thus increasing the network traffic since there are an enormous number of such clients and such links.
  • U.S. Patent Publication Number US20040042398 to Peleg et al., “Method and apparatus for reducing traffic congestion by preventing allocation of the occupied portion of the link capacity and for protecting a switch from congestion by preventing allocation of some of its links,” describes 1) a method for estimating the traffic load between two network nodes over a link of a given capacity and 2) use of such traffic estimates on the links connected to a network switch in a scheme to reduce the chance of congestion through the switch by preventing allocation of portion of the links capacity to one or more clients. While this addresses switch congestion it says nothing about how to more efficiently use multiple access points available to a user communication device.
  • U.S. Patent Publication Number US20040057379 to Chen et al., “Method and apparatus for identifying delay causes in traffic traversing a network,” describes a monitoring scheme which measures the congestion at a given point of a network located on a unique link between a client and a server. Measurement of congestion at a link is of no help in efficient use of multiple access points by a device, and places an additional burden on the network and on computational memory resources of a user device to make use of such data.
  • U.S. Patent Publication Number US20050270985 to Hao et al., “Accelerated per-flow traffic estimation,” describes another network traffic measurement scheme. Additional network traffic is required to distribute the measured data and additional computational memory resources are required to use it. There is no suggestion in this disclosure of how traffic monitoring could be used to support sharing by a client of multiple access points, and other traffic monitoring schemes could more economically do so.
  • U.S. Pat. No. 6,427,114 to Olsson, “Method and means for traffic route control,” describes dynamic traffic control using a road network to minimize delays. Nothing is said about multiple links that are directly connected to a client, and how a client can share these links. U.S. Pat. No. 6,512,760 to Chen, “Alternate wide area network access facility for locally networked computing devices,” describes a method by which a computer with no direct wide area access to the Internet is able to access the Internet indirectly through one or more computers which have access to the web.
  • U.S. Pat. No. 6,574,669 to Weaver, “Method and apparatus for routing traffic within a network utilizing linear optimization,” assumes a network of links with a predetermined capacity for each link and describes a scheme for distributing a given amount of traffic from a source to a destination such that 1) the given throughput of the link is not exceeded and 2) the maximum network capacity is attained from the effective network of paths. In reality the capacity of each link is actually variable since traffic through the network is not constant over time and each link capacity must be monitored and must be reported constantly to the route calculator so as to adjust to the variable capacity of each link. This disclosure provides a routing algorithm but does not say anything about client devices or how a client could efficiently access bandwidth available from multiple links to the network. Furthermore, the disclosed routing algorithm is complex and it would be necessary to wait in order to make use of the link capacity between nodes, doing complex calculations and distributing traffic on the attached links accordingly. However, such calculated data would already be out of date by the time the computations are done.
  • Several prior art references are concerned with SS7 networks. SS7 provides out-of-band signaling and data interfaces between phone switches to reduce congestion in the public switch telephone network. U.S. Patent Publication Number US20040137904 to Gradischnig, “Method and device for routing messages in SS7 networks,” is concerned with source-based routing as opposed to destination-based routing. In Internet networks, packets are routed based on destination address: they are released from the source, routed through the network to the destination according to the routing tables in each node, queued when the outgoing link of a node is busy and are thrown out when they stay too long in the network. U.S. Pat. No. 5,650,998 to Angenot et al., “Method for reassigning traffic in a common channel signaling system (SS7),” is about balancing traffic from one point to another using a set of parallel links in an SS7 telecom network. It describes a method whereby the traffic that used to be carried by a dead link is redistributed among the available links and how it is reassigned to the link when it comes live again. Such link carries a set of point-to-point communication SLS field values and when the link dies, its SLS field values are reassigned to the available live links. U.S. Pat. No. 6,965,598 to Yi describes another signal traffic routing scheme in a telecom network such as SS7. This one selects a link in a link set according to link determination history and link determination data. None of these disclosures indicate how a client can share multiple links, nor do they involve any point-to-point links that are dedicated for the duration of a transmission.
  • SUMMARY OF THE INVENTION
  • It is therefore an object of the present invention to provide Internet applications with access to bandwidth from all available access points, without being limited to a single physical link.
  • Another object of the invention is to distribute among all available links the total traffic needed by all applications running on the client.
  • A further object of the invention is to partition the total client traffic into sub streams to match the different bandwidths offered by each link in order to maximize the overall link capacity, using static or dynamic partitioning.
  • It is also an object of the invention to enhance the browser so that requests for files that are large by nature, such as graphic and video files, are especially identified and assigned to queues corresponding to the larger capacity links.
  • Yet another object of the invention is to provide an improvement that would allow browser and email application to classify their internet requests and dynamically distributes them among queues for each of the accessible points.
  • Another object of the invention is to provide an operating system driver that makes simultaneous use of the all the accessible links and does further statistical distribution of the Internet traffic for all the client applications among these links.
  • It is a further object of the invention to provide a hardware device that incorporates the common functions in each adapter to access multiple access points, thus replacing multiple wireless adapters with a single device and reducing overall hardware cost.
  • It is also an object of the invention to add to nearby routers optimization software that is conscious of the fact that some of the received packets are actually destined to the same client, so that when a given client link is under heavy traffic some packets destined for the client link are rerouted to paths that use other links of that client that are less loaded.
  • In the ideal case, the total bandwidth available to the client should be the sum of the available bandwidth offered by each link. But in reality this is not true because of the statistical nature of the data traffic needed by each application and the randomness of the available bandwidth offered by each link.
  • An aspect of the invention is a method for distributing device communications traffic over multiple access points by identifying a plurality of access points available to the device, instantiating a subset of an OSI communication stack for each identified access point, and directing output requests from the device to the access points so as to make available to the device a total bandwidth that is greater than is available from the respective access points separately. The directing can be based on a round-robin sequencing of the access points, but in another aspect of the invention can be based on assigning priority levels to each of the output requests according to a quality of service requirement of said respective output request. In a further aspect of the invention the output requests can be directed based on monitoring the congestion level of each access point, and assigning higher priority output requests to the least congested access points. Each access point can be served by a separate adapter, or a plurality of access points can be accessed by a multi-point adapter. The invention may be implemented at various points in the OSI communications stack, preferably above the IP level.
  • Note that our approach is different from smart server software (e.g. as promoted by Akamai) where delays are reduced by pre-pushing, but where the bandwidth is unchanged and there is no intended sharing of a physical link.
  • BRIEF DESCRIPTION OF THE DRAWINGS
  • The foregoing and other objects, aspects and advantages will be better understood from the following detailed description of a preferred embodiment of the invention with reference to the drawings, in which:
  • FIG. 1 is a schematic diagram showing the invention implemented as a module layered above TCP/UDP level of the OSI model.
  • FIG. 2 is a schematic diagram showing the invention implemented as a module layered above the IP level of the OSI model.
  • FIG. 3 is a schematic diagram showing the invention implemented as a module layered above the device driver level of the OSI model.
  • DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT OF THE INVENTION
  • In today's client platforms, applications are designed to use one and only one communication adapter to connect to the Internet. This design limitation is based partly on historical reasons. It is also due in part to the fact that the concurrent use of multiple communication adapters requires additional software and, consequently, is more complex. Thus, the bandwidth offered by the other available adapters is not exploited.
  • The present invention describes an intelligent communication agent that distributes incoming and outgoing traffic from the internet among multiple adapters. This agent also determines the optimal communication channel to use for Internet requests by matching the quality of service (QOS) requirement of each request to the communication channel that currently supports such requirements more closely.
  • The invention can be applied to any level of the communication stack. The modules that form the Open System Interconnection (OSI) communication stack are independent of each other. Any portion of the stack above the module implementing the invention would have a single instance and therefore a single set of I/O queues while the portion below the module implemented by the invention would have multiple instances and therefore multiple I/O queues, one set for each adapter. For the returned replies, packets from multiple output queues from lower layer instances would be passed by the invention's component which would simply feed them on to the single upper layer input queue for reassembly into application messages.
  • This is illustrated in FIGS. 1, 2 and 3, to which we now turn. FIG. 1 shows the communication agent 10 of the invention implemented as a module layered above TCP/UDP level 140 of the OSI model, that is, with the application levels (application services level 120 and applications level 110) of the communication stack. For each adapter 170 the communication agent 10 instantiates a subset of the OSI communication stack from the TCP level 140 all the way down to the data link control (DLC) so that each stack is bound to a different MAC address. By this we mean that N sets of data streams are created, each bound to a different adapter 170 and where N is the number of adapters.
  • The communications agent 10 is a software module which sits directly above the TCP communication stack 140 but below the standard communication service applications 120 such as HTTP, FTP, Telnet, etc. Note that this component 10 is sandwiched between components with well documented APIs. The added module 10 consists basically of a giant software switch 20. In the case of outgoing requests, the switch 20 manipulates the output queues of the application levels 100 above and redirects them to the TCP (or UDP) queues 140 of the communication stack 130 associated with each adapter 170. For incoming replies, the switch simply takes the output queue elements of the TCP instances 140 below and feeds them to the input queues of the application levels 100 (i.e. applications 110 via application services 120) above.
  • At first the packets are assigned to the input queues of various stacks in a round robin manner so that we need not have any information on the congestion level of each adapter 170 or QOS requirements of the applications 110 and no further processing is needed. In a more sophisticated implementation of the invention, the communication agent 10 includes a component 30 for assigning priority levels to packets from applications 110 according to their QOS requirements. Such QOS requirements are achieved by making efficient use of the adapter congestion level and other properties of each adapter 170. For some application services 120 such as HTTP, the packet header may be examined so that the packet can be classified and assigned to a communication stack queue according to its specific QOS.
  • In a further implementation of the invention there is added to the communication agent 10 a software module 40 that constantly monitors the congestion level of each adapter 170 and other adapter properties. This is done not only for the upload traffic but also for the return download traffic. A link congestion table maintained by this process is consulted and higher priority requests are assigned to the least congested queues. Other assignment criteria are also used taking into account other properties of the link. Software locks may also be used. On download trips, replies or downloaded traffic are passed from the various communication stacks to the software switch 20 which routes back such replies to the proper application service 120.
  • Note that the invention may be applied below the TCP/UDP level 140, as shown in FIG. 2, or below the IP level 150, as shown in FIG. 3. A similar software switching mechanism as described above applies to these additional configurations. In all of these implementations software switch 20 simply selects the adapter to which the current outgoing packet should be assigned; the selection algorithm is independent of the position of the software switch 20 in the stack. The simplest selection criterion is round robin, where no other action is performed (except in the implementation illustrated in FIG. 3, as described below).
  • More complex selection algorithms may involve a variety of criteria based on QOS data (as a general term which can represent static or dynamic information) and/or traffic congestion estimates on the links: software switch 20 may examine the input packet headers for relevant information to determine the QOS and traffic congestion estimates to use. For example: with downloading applications such as web requests, the inbound link congestion estimate of an adapter is of interest, whereas for uploading applications such as real time collaboration, the outbound link congestion is the one of interest. This does not change with the position of software switch 20 in the stack. However, since software switch 20 may have an upper layer which examines the packet headers for the relevant information, this layer may be more complex as software switch 20 is moved down the stack since there are potentially more headers to look at.
  • For incoming packets from the network, software switch 20 simply passes the packet to the upper layer (or application/application service as shown in the implementation illustrated in FIG. 1) and no processing is involved. Here, again, the functionality of software switch 20 is the same regardless of its position in the stack.
  • However, there are a number of functions which do depend upon where the software switch 20 is located in the stack. Software switch 20 does not modify the outgoing packet except in the implementation shown in FIG. 3: the source address field in the IP packet may be modified in order to match that of the selected adapter. This task is performed even for the simplest selection algorithm, round robin assignment. Further, note that device characteristics may be considered when deciding the preferred position of the software switch 20 in the stack: the higher in the stack software switch 20 is placed, the faster it is able to service requests, but the more memory used. So the implementation shown in FIG. 3 may be preferred for memory constrained devices and the implementation shown in FIG. 1 may be preferred for devices where speed is paramount and memory is no concern. Also, note that in the implementation shown in FIG. 1, there is no optional nearby router optimization because the TCP layer performs its own message break up and reassembly.
  • A single Wi-Fi adapter cannot connect to several access points using Time Division Multiple Access (TDMA). With TDMA, used by GSM cellular, the radio band is divided into fixed time slots assigned to different signals so that each adapter waits for reserved time slots to transmit and receive. With Wi-Fi, which uses spread-spectrum technology, adapters transmit and receive continuously over the entire band. So one can think of the suggested TDMA idea as time sharing of one adapter, the very opposite of the goal of increased bandwidth via concurrent use of multiple adapters in order to increase the total communication capacity. Only when multiple adapters transmit and/or receive concurrently, will the effective bandwidth increase and the inherent delays be reduced.
  • Note also that, even in the vicinity of a single Wi-Fi access point, a client with multiple Wi-Fi adapters increases its net bandwidth because it gets a larger fraction of the total available bandwidth than it would with a single adapter, assuming there are other users simultaneously vying for the access point traffic.
  • It is desirable to use multiple connections instead of the “best” connection. In the world of rapidly changing traffic loads and rapidly developing technology, the notion of “best” is ephemeral: the load on a link changes instantly by the mere fact that you start using it. Second, with multiple adapters, you still have access to the “best” and then some. Lastly, in a hot spot, there may not be one, but multiple “best” connections at any given time.
  • The invention actually does assign the “best” connection, but it does so at the request level, dynamically and according to a flexible set of “best” QOS selection criteria. Plus, the communication agent can optionally be configured to use only a subset of the total number of adapters. Note that in a case where there are more access points than adapters, we periodically update the pool of the n “best” access points.
  • Note that the preferred place for the invention in the communication stack is as illustrated in FIG. 2, directly below TCP/UDP 140 and above IP 150, because the IP layer sets the Internet message source address and the communication agent 10 basically selects the source address to use for the message by assigning it to a given adapter.
  • It makes sense to pack multiple Wi-Fi adapters into a single form factor because transistors are essentially free and consequently by sharing hardware resources such an adapter becomes cost effective.
  • The present invention can make use of any Internet bandwidth available to the device, be it via Wi-Fi, Ethernet, wideband cellular, WiMax etc. The present invention can also be used to increase the outgoing bandwidth from the device to the Internet in cases such as, but not limited to, real time web collaboration or peer-to-peer applications.
  • We will now use several examples to help explain operation of the invention. First, consider the example of several applications running at the same time, for example Email and two instances of the browser. Also assume these applications run on a laptop that in connected to a LAN and also has two Wi-Fi access points in the vicinity. Lets call these three links A, B and C.
  • As a first stage in explaining the invention, these links can be allocated statically. We configure one instance of the browser to connect through link A, the other browser instance is configured to use link B and the Email application is configured to use link C. Each browser instance uses its link to fetch a new page. All child requests of a given page inherit the page's link and use such link to connect and receive data for that given page.
  • Now consider a dynamic link allocation. We also modify the browser's parser so that when a requested page is parsed, the child request parts are fetched using different available links. Link queue allocation could be round robin at first. Later we can investigate more sophisticated assignment schemes taking in account each request's bandwidth requirement, for example, more bandwidth is required for a graphic file. We also patch the Email application similarly.
  • In a third stage of demonstrating how the invention operates we modify the OS driver to the link to service multiple adapter cards simultaneously, taking into account that such a driver must take its data from multiple queues used by multiple applications. So we write a driver that is conscious of the presence of multiple adapter cards, each connected to a different access point. Such a driver takes the Internet requests from the browser and the other Internet applications and dynamically distributes them into the output queues feeding the different links. Again, this can be done in a round robin fashion or by giving priority to the output queue of the link with the least load, assigning the list of all the Internet requests on a first arrived first served basis or some other algorithm.
  • In a fourth stage of demonstration, the driver further classifies the Internet requests according to the expected size, source, and net address groupings. These classifications are used to fill multiple Internet queues. We also assign priority levels to each queue. The OS driver is made to account for these properties in the distribution of the requests among the several output queues assigned to each link.
  • Note that in the general sense, a similar environment can be applied to PDAs or wireless phones accessing the Internet through the cellular phone network where the access points correspond to the accessible cellular band from a given location.
  • A second example used to explain how the invention works is a comparison with an existing technology, and in particular the affinity based routing method described in U.S. Pat. No. 6,424,992 to Devarakonda et al. (hereafter, “affinity routing disclosure”). In the affinity routing disclosure a client sends its requests to a specific router encapsulating a server cluster. The router consults its table for each client request and redirects it to a server node according to an affinity factor. This affinity is based on previous routing history and optionally server load. The affinity table is updated by the server nodes based on their load and availability. The router may itself be a server. The routing table is consulted for incoming client requests. The router is near the destination server, not near the client.
  • IP affinity switching to route requests to servers has some similarities to the method of the present invention. However, upon close examination there are significant differences. In router optimization according to the present invention the routers involved are limited to the ones in the vicinity of the client access points to the network. The routing table is consulted on return paths from the Internet. It contains for each client domain name the list of its IP addresses, and optionally the upload and download congestion factors of each link. Traffic is rerouted according to congestion factors. These congestion factors, as we mentioned earlier, change extremely fast, unlike the affinity factors in the affinity routing disclosure, so they are not based on past history. They are maintained by the client and/or the router and not by a server. Furthermore the table is consulted on the return trip from the server. Also note that each router table has a limited number of records, one for each client in the vicinity of the router, whereas in the affinity routing disclosure the number of entries in the table can become very large.
  • Now, let's compare the affinity routing disclosure to the application level implementation of the present invention. This implementation is described above in connection with FIG. 1. Note that in this implementation of the present invention, there is no router optimization. In the present invention, we are dynamically mapping internet requests to communication channels, unlike the affinity routing disclosure, which uses a fixed policy to assign requests to servers, based on load and/or history. The present invention matches requests QOS characteristics to current communication links characteristics.
  • While the invention has been described in terms of preferred embodiments, those skilled in the art will recognize that the invention can be practiced with modification within the spirit and scope of the appended claims.

Claims (20)

  1. 1. A method for distributing device communications traffic over multiple access points, comprising:
    identifying a plurality of access points available to the device;
    instantiating a subset of an OSI communication stack for each identified access point; and
    directing output requests from the device to one of said plurality of access points so as to make available to the device a total bandwidth that is greater than is available from the respective access points separately.
  2. 2. A method as in claim 1, wherein said directing is based on a round-robin sequence of said access points.
  3. 3. A method as in claim 1, further comprising assigning priority levels to each of said output requests according to a quality of service requirement of said respective output request, wherein said directing is based on said quality of service requirement.
  4. 4. A method as in claim 3, further comprising monitoring a congestion level of each access point, and wherein said directing is based on the congestion level.
  5. 5. A method as in claim 4, wherein said directing assigns higher priority output requests to least congested access points.
  6. 6. A method as in claim 1, wherein each access point is accessed by a separate adapter in said device.
  7. 7. A method as in claim 1, wherein each of said plurality of access points is accessed by a multi-point adapter and said device is one of i) a personal digital assistant or ii) a cellular phone operated on a cellular network.
  8. 8. A method as in claim 1, wherein said subset of an OSI communication stack extends from a TCP/UDP level to a data link control.
  9. 9. A method as in claim 1, wherein said subset of an OSI communication stack extends from an IP level to a data link control.
  10. 10. A method as in claim 1, wherein said subset of an OSI communication stack extends from a network driver to a data link control.
  11. 11. A system for distributing device communications traffic over multiple access points, comprising:
    means for identifying a plurality of access points available to the device;
    means for instantiating a subset of an OSI communication stack for each identified access point; and
    means for directing output requests from the device to one of said plurality of access points so as to make available to the device a total bandwidth that is greater than is available from the respective access points separately.
  12. 12. A system as in claim 11, wherein said directing means is based on a round-robin sequence of said access points.
  13. 13. A system as in claim 11, further comprising means for assigning priority levels to each of said output requests according to a quality of service requirement of said respective output request, wherein said directing means uses said quality of service requirement.
  14. 14. A system as in claim 13, further comprising means for monitoring a congestion level of each access point, and wherein said directing means uses the congestion level.
  15. 15. A system as in claim 14, wherein said directing means assigns higher priority output requests to least congested access points.
  16. 16. A system as in claim 11, wherein each access point is accessed by a separate adapter in said device.
  17. 17. A system as in claim 11, wherein each of said plurality of access points is accessed by a multi-point adapter and said device is a personal digital assistant.
  18. 18. A system as in claim 11, wherein said subset of an OSI communication stack extends from a TCP/UDP level to a data link control.
  19. 19. A computer implemented system for distributing device communications traffic over multiple access points, comprising:
    first computer code for identifying a plurality of access points available to the device;
    second computer code for instantiating a subset of an OSI communication stack for each identified access point; and
    third computer code for directing output requests from the device to one of said plurality of access points so as to make available to the device a total bandwidth that is greater than is available from the respective access points separately.
  20. 20. A computer implemented system as in claim 19, further comprising fourth computer code for assigning priority levels to each of said output requests according to a quality of service requirement of said respective output request, wherein said third computer code for directing is based on said quality of service requirement.
US11433359 2006-05-15 2006-05-15 Increasing link capacity via traffic distribution over multiple WI-FI access points Abandoned US20070263616A1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US11433359 US20070263616A1 (en) 2006-05-15 2006-05-15 Increasing link capacity via traffic distribution over multiple WI-FI access points

Applications Claiming Priority (5)

Application Number Priority Date Filing Date Title
US11433359 US20070263616A1 (en) 2006-05-15 2006-05-15 Increasing link capacity via traffic distribution over multiple WI-FI access points
PCT/US2007/065823 WO2007133862A3 (en) 2006-05-15 2007-04-03 Increasing link capacity via traffic distribution over multiple wi-fi access points
CA 2652080 CA2652080C (en) 2006-05-15 2007-04-03 Increasing link capacity via traffic distribution over multiple wi-fi access points
CN 200780017965 CN101449527A (en) 2006-05-15 2007-04-03 Increasing link capacity via traffic distribution over multiple Wi-Fi access points
US12052319 US8169900B2 (en) 2006-05-15 2008-03-20 Increasing link capacity via traffic distribution over multiple Wi-Fi access points

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US12052319 Continuation US8169900B2 (en) 2006-05-15 2008-03-20 Increasing link capacity via traffic distribution over multiple Wi-Fi access points

Publications (1)

Publication Number Publication Date
US20070263616A1 true true US20070263616A1 (en) 2007-11-15

Family

ID=38685039

Family Applications (2)

Application Number Title Priority Date Filing Date
US11433359 Abandoned US20070263616A1 (en) 2006-05-15 2006-05-15 Increasing link capacity via traffic distribution over multiple WI-FI access points
US12052319 Expired - Fee Related US8169900B2 (en) 2006-05-15 2008-03-20 Increasing link capacity via traffic distribution over multiple Wi-Fi access points

Family Applications After (1)

Application Number Title Priority Date Filing Date
US12052319 Expired - Fee Related US8169900B2 (en) 2006-05-15 2008-03-20 Increasing link capacity via traffic distribution over multiple Wi-Fi access points

Country Status (4)

Country Link
US (2) US20070263616A1 (en)
CN (1) CN101449527A (en)
CA (1) CA2652080C (en)
WO (1) WO2007133862A3 (en)

Cited By (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090028151A1 (en) * 2007-07-27 2009-01-29 Schmidt Brian K Packet level prioritization in interconnection networks
US20090265414A1 (en) * 2007-11-29 2009-10-22 Bryan David A Mechanisms for transparently converting client-server software agents to peer-to-peer software agents
US7756134B2 (en) 2006-05-02 2010-07-13 Harris Corporation Systems and methods for close queuing to support quality of service
US7769028B2 (en) 2006-06-21 2010-08-03 Harris Corporation Systems and methods for adaptive throughput management for event-driven message-based data
US7856012B2 (en) 2006-06-16 2010-12-21 Harris Corporation System and methods for generic data transparent rules to support quality of service
US7894509B2 (en) 2006-05-18 2011-02-22 Harris Corporation Method and system for functional redundancy based quality of service
US7916626B2 (en) 2006-06-19 2011-03-29 Harris Corporation Method and system for fault-tolerant quality of service
US20110167165A1 (en) * 2007-11-29 2011-07-07 Bryan David A Mechanisms for transparently converting client-server software agents to peer-to-peer software agents
US7990860B2 (en) 2006-06-16 2011-08-02 Harris Corporation Method and system for rule-based sequencing for QoS
US8064464B2 (en) 2006-06-16 2011-11-22 Harris Corporation Method and system for inbound content-based QoS
US8300653B2 (en) 2006-07-31 2012-10-30 Harris Corporation Systems and methods for assured communications with quality of service
US8516153B2 (en) 2006-06-16 2013-08-20 Harris Corporation Method and system for network-independent QoS
US8730981B2 (en) 2006-06-20 2014-05-20 Harris Corporation Method and system for compression based quality of service
US20160164801A1 (en) * 2014-12-08 2016-06-09 Samsung Electronics Co., Ltd. Method and apparatus for setting communicating session
US20170280361A1 (en) * 2015-09-22 2017-09-28 Veniam, Inc. Systems and methods for managing connectivity between access points and hotspots in a network of moving things, for example including autonomous vehicles.

Families Citing this family (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8719399B2 (en) * 2005-04-07 2014-05-06 Opanga Networks, Inc. Adaptive file delivery with link profiling system and method
US8909807B2 (en) 2005-04-07 2014-12-09 Opanga Networks, Inc. System and method for progressive download using surplus network capacity
US8589508B2 (en) * 2005-04-07 2013-11-19 Opanga Networks, Inc. System and method for flow control in an adaptive file delivery system
US7500010B2 (en) 2005-04-07 2009-03-03 Jeffrey Paul Harrang Adaptive file delivery system and method
US9065595B2 (en) 2005-04-07 2015-06-23 Opanga Networks, Inc. System and method for peak flow detection in a communication network
WO2008089277A1 (en) * 2007-01-16 2008-07-24 Mediacast, Llc Wireless data delivery management system and method
WO2010017205A3 (en) * 2008-08-04 2010-04-22 David Gibbons Systems and methods for video bookmarking
EP2350962A4 (en) * 2008-09-18 2013-08-21 Opanga Networks Inc Systems and methods for automatic detection and coordinated delivery of burdensome media content
US8589579B2 (en) * 2008-10-08 2013-11-19 Citrix Systems, Inc. Systems and methods for real-time endpoint application flow control with network structure component
JP2012508424A (en) * 2008-11-07 2012-04-05 オパンガ ネットワークス インコーポレイテッド Portable data storage device to start the data transfer by using a host device
US20100131385A1 (en) * 2008-11-25 2010-05-27 Opanga Networks, Llc Systems and methods for distribution of digital media content utilizing viral marketing over social networks
US8886790B2 (en) * 2009-08-19 2014-11-11 Opanga Networks, Inc. Systems and methods for optimizing channel resources by coordinating data transfers based on data type and traffic
EP2468067A4 (en) 2009-08-19 2015-09-09 Opanga Networks Inc Optimizing media content delivery based on user equipment determined resource metrics
EP2468019A4 (en) * 2009-08-20 2013-10-23 Opanga Networks Inc Broadcasting content using surplus network capacity
US8495196B2 (en) 2010-03-22 2013-07-23 Opanga Networks, Inc. Systems and methods for aligning media content delivery sessions with historical network usage
US9603033B2 (en) 2011-12-31 2017-03-21 International Business Machines Corporation Online and distributed optimization framework for wireless analytics
US20130305344A1 (en) * 2012-05-14 2013-11-14 Alcatel-Lucent India Limited Enterprise network services over distributed clouds
CN103686756B (en) * 2012-09-17 2016-12-21 中国科学院沈阳自动化研究所 tdma access device and access method based on multi-access point
CN103297432B (en) * 2013-05-28 2016-10-26 南京信息工程大学 An artificial weather modification network access server running method
US9621400B2 (en) * 2013-05-31 2017-04-11 Avago Technologies General Ip (Singapore) Pte. Ltd. Systems and methods for transmitting content
CN103414654B (en) * 2013-08-26 2017-03-29 深圳市德传技术有限公司 One implementation, the bandwidth of the shunt system and method

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5519709A (en) * 1994-05-04 1996-05-21 Hewlett-Packard Company Two priority fair distribution round robin protocol for a network having cascaded hubs
US20030076837A1 (en) * 2001-10-23 2003-04-24 Whitehill Eric A. System and method for providing a congestion optimized address resolution protocol for wireless Ad-Hoc Networks
US20050271048A1 (en) * 2004-06-04 2005-12-08 Liam Casey Selective internet priority service
US20060050736A1 (en) * 2004-09-08 2006-03-09 Alcatel Rural broadband hybrid satellite/terrestrial solution

Family Cites Families (26)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0681405A1 (en) * 1994-05-06 1995-11-08 International Business Machines Corporation System for diverting signalling link traffic in a common channel signalling system
US6396816B1 (en) * 1994-12-20 2002-05-28 Intel Corporation Method and apparatus for multiple applications on a single ISDN line
DE69924619T2 (en) * 1998-08-07 2006-02-02 Dinbis Ab A method and device for monitoring a traffic route
US6574669B1 (en) * 1998-08-31 2003-06-03 Nortel Networks Limited Method and apparatus for routing traffic within a network utilizing linear optimization
KR100320907B1 (en) * 1999-05-06 2002-02-04 서평원 signal traffic routing method of NO.7 signaling
CN1111806C (en) * 1999-06-30 2003-06-18 北京爱必得电脑有限责任公司 Architecture of video play-on-demand system
US6512760B1 (en) * 1999-08-09 2003-01-28 Avinta Communications, Inc Alternate wide area network access facility for locally networked computing devices
US20040136379A1 (en) * 2001-03-13 2004-07-15 Liao Raymond R Method and apparatus for allocation of resources
US7000026B2 (en) * 2000-12-22 2006-02-14 Nortel Networks Limited Multi-channel sharing in a high-capacity network
US20020080010A1 (en) * 2000-12-22 2002-06-27 George Zhang Power line communications network device for DC powered computer
US6724874B2 (en) * 2001-01-17 2004-04-20 Sbc Technology Resources, Inc. Outgoing call screening
US20020167954A1 (en) * 2001-05-11 2002-11-14 P-Com, Inc. Point-to-multipoint access network integrated with a backbone network
DE60234665D1 (en) * 2001-06-21 2010-01-21 Sk Telecom Co Ltd A method for the determination of the path in a network with multi-protocol label switching
US20030163593A1 (en) * 2002-02-25 2003-08-28 William March Rice University Method and system for implementing a fair, high-performance protocol for resilient packet ring networks
US20040042398A1 (en) * 2002-02-28 2004-03-04 Seriqa Networks Method and apparatus for reducing traffic congestion by preventing allocation of the occupied portion of the link capacity and for protecting a switch from congestion by preventing allocation on some of its links
US20040022222A1 (en) * 2002-07-31 2004-02-05 Allister Clisham Wireless metropolitan area network system and method
DE50203898D1 (en) * 2002-09-03 2005-09-15 Siemens Ag Method and apparatus for message routing in the SS7 network
US7221649B2 (en) * 2002-09-20 2007-05-22 Lucent Technologies Inc. Method and apparatus for identifying delay causes in traffic traversing a network
US20040203828A1 (en) * 2002-12-06 2004-10-14 Vinod Mirchandani Selective channel assignment in a wireless data communications network
CN100411447C (en) * 2003-11-06 2008-08-13 华为技术有限公司 Configurating method and apparatus for channel band width in telecommunicating system
US7460487B2 (en) * 2004-06-04 2008-12-02 Lucent Technologies Inc. Accelerated per-flow traffic estimation
US7512705B2 (en) * 2004-12-01 2009-03-31 Hewlett-Packard Development Company, L.P. Truncating data units
US7480264B1 (en) * 2005-02-10 2009-01-20 Sonicwall, Inc. Centralized wireless LAN load balancing
EP1701487A1 (en) * 2005-03-07 2006-09-13 France Telecom Bandwidth adaptation according to network load
US7502348B2 (en) * 2005-04-01 2009-03-10 Toshiba America Research, Inc. Silent proactive handoff
CN100416506C (en) 2006-10-27 2008-09-03 北京金山软件有限公司;北京金山数字娱乐科技有限公司 Method and system for distributing data of net surfing server based on multiple nuclear CPU framework

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5519709A (en) * 1994-05-04 1996-05-21 Hewlett-Packard Company Two priority fair distribution round robin protocol for a network having cascaded hubs
US20030076837A1 (en) * 2001-10-23 2003-04-24 Whitehill Eric A. System and method for providing a congestion optimized address resolution protocol for wireless Ad-Hoc Networks
US20050271048A1 (en) * 2004-06-04 2005-12-08 Liam Casey Selective internet priority service
US20060050736A1 (en) * 2004-09-08 2006-03-09 Alcatel Rural broadband hybrid satellite/terrestrial solution

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7756134B2 (en) 2006-05-02 2010-07-13 Harris Corporation Systems and methods for close queuing to support quality of service
US7894509B2 (en) 2006-05-18 2011-02-22 Harris Corporation Method and system for functional redundancy based quality of service
US7990860B2 (en) 2006-06-16 2011-08-02 Harris Corporation Method and system for rule-based sequencing for QoS
US7856012B2 (en) 2006-06-16 2010-12-21 Harris Corporation System and methods for generic data transparent rules to support quality of service
US8516153B2 (en) 2006-06-16 2013-08-20 Harris Corporation Method and system for network-independent QoS
US8064464B2 (en) 2006-06-16 2011-11-22 Harris Corporation Method and system for inbound content-based QoS
US7916626B2 (en) 2006-06-19 2011-03-29 Harris Corporation Method and system for fault-tolerant quality of service
US8730981B2 (en) 2006-06-20 2014-05-20 Harris Corporation Method and system for compression based quality of service
US7769028B2 (en) 2006-06-21 2010-08-03 Harris Corporation Systems and methods for adaptive throughput management for event-driven message-based data
US8300653B2 (en) 2006-07-31 2012-10-30 Harris Corporation Systems and methods for assured communications with quality of service
US7911956B2 (en) * 2007-07-27 2011-03-22 Silicon Image, Inc. Packet level prioritization in interconnection networks
US20090028151A1 (en) * 2007-07-27 2009-01-29 Schmidt Brian K Packet level prioritization in interconnection networks
US20110167165A1 (en) * 2007-11-29 2011-07-07 Bryan David A Mechanisms for transparently converting client-server software agents to peer-to-peer software agents
US20090265414A1 (en) * 2007-11-29 2009-10-22 Bryan David A Mechanisms for transparently converting client-server software agents to peer-to-peer software agents
US9332068B2 (en) 2007-11-29 2016-05-03 Ooma, Inc. Mechanisms for transparently converting client-server software agents to peer-to-peer software agents
US20160164801A1 (en) * 2014-12-08 2016-06-09 Samsung Electronics Co., Ltd. Method and apparatus for setting communicating session
US20170280361A1 (en) * 2015-09-22 2017-09-28 Veniam, Inc. Systems and methods for managing connectivity between access points and hotspots in a network of moving things, for example including autonomous vehicles.
US9913175B2 (en) * 2015-09-22 2018-03-06 Veniam, Inc. Systems and methods for managing connectivity between access points and hotspots in a network of moving things, for example including autonomous vehicles

Also Published As

Publication number Publication date Type
CA2652080A1 (en) 2007-11-22 application
CN101449527A (en) 2009-06-03 application
WO2007133862A3 (en) 2008-04-10 application
US8169900B2 (en) 2012-05-01 grant
CA2652080C (en) 2014-12-02 grant
US20080165693A1 (en) 2008-07-10 application
WO2007133862A2 (en) 2007-11-22 application

Similar Documents

Publication Publication Date Title
US6760775B1 (en) System, method and apparatus for network service load and reliability management
US6389448B1 (en) System and method for load balancing
US7477657B1 (en) Aggregating end-to-end QoS signaled packet flows through label switched paths
US7010611B1 (en) Bandwidth management system with multiple processing engines
Suter et al. Buffer management schemes for supporting TCP in gigabit routers with per-flow queueing
US6865185B1 (en) Method and system for queuing traffic in a wireless communications network
US20030120705A1 (en) Method and apparatus for providing multiple data class differentiation with priorities using a single scheduling structure
US6940814B1 (en) System and method for a quality of service in a multi-layer network element
US6330602B1 (en) Scaleable web server and method of efficiently managing multiple servers
US6587469B1 (en) Telecommunications system
US5629928A (en) Dynamic fair queuing to support best effort traffic in an ATM network
US20040143662A1 (en) Load balancing for a server farm
US6728748B1 (en) Method and apparatus for policy based class of service and adaptive service level management within the context of an internet and intranet
US20040199664A1 (en) Method and system for improving a route along which data is sent using an ip protocol in a data communications network
US6977932B1 (en) System and method for network tunneling utilizing micro-flow state information
US9270598B1 (en) Congestion control using congestion prefix information in a named data networking environment
US20030061338A1 (en) System for multi-layer broadband provisioning in computer networks
US20130212279A1 (en) Resource Allocation Mechanism
US7953885B1 (en) Method and apparatus to apply aggregate access control list/quality of service features using a redirect cause
US20120075991A1 (en) Network system, control method thereof and controller
US5918021A (en) System and method for dynamic distribution of data packets through multiple channels
US20020075803A1 (en) Method and apparatus for dynamic optimization of a multi-service access device
US5687167A (en) Method for preempting connections in high speed packet switching networks
EP1035751A2 (en) Adaptive routing system and method for Qos packet networks
US6473434B1 (en) Scaleable and robust solution for reducing complexity of resource identifier distribution in a large network processor-based system

Legal Events

Date Code Title Description
AS Assignment

Owner name: INTERNATIONAL BUSINESS MACHINES CORPORATION, NEW Y

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:CASTRO, PAUL CHRISTESEN;GIRAUD, FREDERIQUE ANNE;YEH, DANNY LO-TIEN;REEL/FRAME:017842/0073;SIGNING DATES FROM 20060512 TO 20060515