JP2004514355A - Multiple service sub-throws in cable modem service flow - Google Patents

Abstract

A disclosure of a method for subdividing a first network data flow into a plurality of sub-flows. Each subflow is assigned a special service class or priority value. Tokens are assigned, such as by receiving the highest priority packet as a higher transmission priority, to coordinate data transmission over the network from the first flow to each subflow.

Description

[0001]
Field of the invention
The present invention relates generally to communication systems, and more particularly to communication systems in which a service data flow is subdivided into a plurality of subflows.
[0002]
Background of the Invention
Internet access using telephone modems is currently being provided at speeds up to 56 Kbps. Telephone-based modems modulate and demodulate data signals for transmission over the voice band of the telephone network. In contrast, cable modems use a cable television system to provide access to the Internet and other external networks. This provides the cable modem with a higher bandwidth and operates at a higher data rate than the telephone system. A cable modem allows a connection between a user's computer and other communication devices and a cable system headend, for example, from a headend to an external network through a T1 Transmission Line or the like. In cable networks, data transferred from the network headend to users and subscribers is called downstream data; data transferred from the user to the network headend is called upstream data.
[0003]
A typical state of the art bidirectional cable system is illustrated in block diagram form in FIG. Current technology cable systems include a head-end device 101, an optical coaxial hybrid (HFC) cable plant 103, a plurality of cable modems 105, 106 (two shown), and associated cable modems 106, 117 through communication links 116, 117. It comprises an associated number of subscriber communication devices 107, 108 (two shown) connected to 107. The subscriber's communication devices 107, 108 may comprise a computer, a television, a telephone, or the like. As is well known to those skilled in the art, the headend device 101 includes a processor, a router, a switch, a downstream transmitter, an upstream receiver, a splitter, a combiner, a subscriber database, a network management station, a dynamic host configuration protocol (DHCP), a trivial. It consists of a file transfer protocol (TFTP) service, call agent, media gateway, and billing system. The HCF cable plant 103 comprises fiber optic cables, coaxial cables, fiber / coaxial nodes, amplifiers, filters, and taps, which transmit from the headend device 101 to the cable modems 105, 106 via a shared downstream channel 110 and the cable modems 105, 106. It supports transmission from 106 to headend device 101 over shared upstream channel 112. The program signal is input to the headend device 101 for broadcasting to the subscriber through the HFC cable plant 103.
[0004]
The downstream channel 110 and the upstream channel 112 use respective transmission protocols for communication information. As a common example, the modulation scheme used to transmit information over the downstream channel 110 (eg, 64-array quadrature amplitude modulation (QAM)) is the modulation used to transmit information over the upstream channel 112 (eg, , Higher than Differential Four-Phase Phase Modulation (DQPSK) or 16-array QAM), and as a result, the downstream transmission rate is faster than the upstream transmission rate. Cable systems in which the upstream transmission rate is lower than the downstream transmission rate are commonly referred to as "asynchronous" systems. Cable systems in which the upstream transmission rate is similar to the downstream transmission rate are called "synchronous" systems.
[0005]
In addition to the specific type of modulation used on each channel 110, 112, the shared nature of each channel 110, 112 introduces other protocol requirements. For example, if the downstream channel 110 is shared, the downstream protocol includes address information, and each cable modem 105, 106 monitors the downstream channel 110 for information packets destined there. On the other hand, information directed to specific cable modems 105 and 106 (connected communication devices 107 and 108, and broadcast messages (such as broadcast messages) to all cable modems 105 and 106 (and connected communication devices 107 and 108)). The packets are processed by cable modems 105, 106 and forwarded to connected subscriber communication devices 107, 108 (eg, telephones, personal computers, other terminating devices) as needed.The upstream channel is shared. In some cases, upstream channel access protocols are used to reduce the likelihood of collision of communication information resulting from cable modems 105, 106. Numerous access protocols are used for upstream channel access. Among them, there are ALOHA, slotted ALOHA, code division multiple access (CDMA), time division multiple access (TDMA), collision detection type TDMA (TDMA \ With \ Collision \ Detect), and carrier detection. Well-known protocols such as multiple access are included.
[0006]
Some bidirectional cable systems comply and use the upstream and downstream channel protocols defined by the recently published Data over Cable System Interface (DOCSIS) version 1.0.
[0007]
The details of DOCIS are incorporated herein. The upstream protocol defined by the DOCSIS standard is controlled in timing by the headend device 101 (“Cable Modem Termination System” (CMTS) in the DOCSIS standard) and via a timestamp synchronization message transmitted over the downstream channel 110. Connect to cable modems 105 and 106. Therefore, in order to generate upstream communication with high accuracy (high quality @ manner) in turn, each cable before transmission of the information provided by the subscriber's communication device 107, 108 at the modem 105, 106 begins. The time references in the modems 105 and 106 and similar time references in the headend device 101 need to be substantially synchronized. Otherwise, transmissions from modem 105 may conflict with transmissions from other modems 106.
[0008]
The headend device 101 is typically connected to an external network 114, such as a public switched telephone network, via a suitable communication link 119, such as a fiber optic distribution data interface (FDDI) link or a 100 base T Ethernet link. It is connected to a wide area packet network such as the Internet. Thus, the two-way cable system communicates between the subscriber's communication devices 107 and 108, similar devices not included in FIG. 1, Internet servers, computer networks, and other devices represented by the external network 114. Provides connection connectivity.
[0009]
Also, the headend device 101 receives a program signal for broadcast to subscribers (via satellite downlink, terrestrial microwave, wire).
[0010]
Subscriber data is 6MH_zSent through the channel. 6MH_zChannel is the size of the band allocated for cable television channels for broadcasting television signals for all subscribers. On the subscriber side, the program signal is received by the set top box (FIG. 2) while the downstream data is received by the cable modems 105 and 106, respectively. Many upstream and downstream channels in a cable modem system are processed based on service area, number of users, data rate contracted to each user, and available bandwidth.
[0011]
FIG. 2 is a block diagram of the cable system components at the subscriber and user side and is associated with the communication device 108. At the subscriber's premises, a splitter 134 distributes input signals from the cable system headend 101. The television signal is displayed on the television 140 under the control of the set-top box 138. A second output from splitter 134 allows connection to cable modem 105. The downstream signal from the head end device 101 is supplied to an RF (radio frequency) tuner 142. The signal is tuned to the frequency band assigned to the cable modem 105 during the startup and configuration phases of the cable modem 105. As mentioned above, the downstream channel is typically demodulated in demodulator 144 using quadrature amplitude modulation (QAM). The demodulated signal is input to the media access controller 146. The base-end data signal from the media access controller 146 is input to a data / management logic unit that provides overall control of the cable modem 105 and provides data control functions. Communication device 108 is connected to the data / management logic unit of cable modem 105 for receiving data transmitted in the downstream direction and transmitting data in the upstream direction.
[0012]
The upstream data is sent from the communication device 108 through the data / management logic unit, the media controller, and finally to the modulator for modulation. Generally, differential four-phase modulation (DQPSK) or 16QAM modulation is adopted for upstream data. The selected modulation type is described in the setting information supplied to each cable 105. The upstream data passes through the splitter 134 for transmission and is sent out to the headend 101 via the optical coaxial cable plant 103. Eventually, the data reaches the external network 114 as described in FIG.
[0013]
In one implementation, the downstream channel is 6 MH_z64QAM modulation or 256QAM modulation capable of distributing data on the cable channel at 30 to 40 Mbps is adopted. The upstream channel uses QPSK or 16QAM signals with data rates available from 320 Kbps to 10 Mbps. Upstream and downstream data rates are set by the system operator. For example, a cable modem 105 used by business users can be configured to receive and transmit information at a high data rate in both directions. On the other hand, a home user may have a cable modem 105 with upstream data transmission limited to low speeds and set for wide bandwidth access in the downstream direction (hence high data rate) to receive data from the external network 114. . In yet another implementation, the data rate allocation can be time-of-day-sensitive.
[0014]
However, it must be emphasized that cable modems use a specified modulation type and operate at a specified modulation rate (or data rate). The priority of the independent data packets does not affect the set upstream data rate or bandwidth. All modems transmitting on the upstream channel use the same modulation rate and modulation type. However, the actual amount of bandwidth used by a cable modem is limited by the software that controls the modem. As shown above, while the home user is limited to 1 Mbps, the business user is set to 2 Mbps in the upstream direction. However, cable modems supporting business and home users are capable of using at high system rates. The priority of each data packet does not affect the set modulation rate or modulation type.
[0015]
When the cable modem 105 or 107 is turned on, a connection is made to the headend device 101 via the optical coaxial cable plant 103. This connection employs the Internet Protocol (IP), so that data coming from the Internet or the World Wide Web (typically the external network 114) in IP format as received by the headend device 101 is transmitted to the cable modem 105 or Downstream can be advanced to 106. During the process of initializing the cable modem, the modem obtains the IP address, other operational parameters, the server address in the modem configuration file from the Dynamic Host Configuration Protocol (DHCP), and the like. Any DHCP server can respond to the broadcast request and supply the required information. The configuration file consists of various modem configuration parameters, which consist of access control information, assignment of downstream and upstream channels, and a Trivial File Transfer Protocol (TFTP) server from which modem operating software can be downloaded.
[0016]
Packet switch networks, such as the cable system network shown in FIG. 1, are typically operated in a best-effort service (delivery) manner. Unfortunately, Internet protocols using the connectionless best effort delivery model do not guarantee that packets will be delivered in order, on time, etc. All traffic has a similar priority and a similar likelihood of being delivered on time. If the network is crowded, and all traffic has similar potential for delays and drops. Many types of networked applications include mission-critical applications that interface to the World Wide Web, online business critical applications, and multimedia-based applications (such as desktop video conferencing and web-based training). And Voice over Internet Protocol (VoIP). Certain network applications may be sensitive to bandwidth and delay, and thus require special classes of service requests on the network. Some applications require the receiver to deliver in real time, while other applications can tolerate a small delay. In order to deploy real-time applications such as VoIP and video conferencing using Internet protocols, an acceptable level of quality, such as bandwidth latency and demand for fluctuations, needs to be ensured. There must be a way to enable coexisting real-time traffic using traditional data traffic. Adopting the concept of class of service (CoS) requires network managers to provide bandwidth for mission-critical and real-time application traffic by allowing lower critical applications to use the network in a rational and efficient manner. Allows you to guarantee protection from other applications. To this end, network service class policies regulate network resources according to the objectives of the network user. Without these service class parameters, network resources would be quickly consumed by non-essential applications and would not be available to more important ones. A class of service implementation recognizes and delivers high priority traffic to network devices in a predictable manner. When network congestion occurs, the service class mechanism drops or delays lower priority traffic in order to deliver higher priority traffic.
[0017]
Therefore, there is a need to establish a class of service scheme in a packet switched cable modem system and develop a mechanism to implement it, thereby increasing the likelihood of high priority upstream data to its destination on time. Will reach.
[0018]
Detailed Description of the Preferred Embodiment
Before describing the details of a particular method or apparatus for assigning or prioritizing service classes to data flows in a two-way cable modem communication system, the present invention first addresses the steps and apparatus associated therewith. You can see that it is in a novel combination. Accordingly, hardware components and method steps are represented by conventional elements in the drawings, showing only certain details pertinent to the present invention, and the advantages described herein provided by the present invention being readily apparent to those skilled in the art. The present disclosure has been clarified by structural details that can be understood by those skilled in the art.
[0019]
The DOCSIS 1.0 protocol allows a cable modem, such as cable modems 105, 106, to set up multiple service flows for carrying data in the upstream direction. However, it is impossible for cable operators to make effective use of this property. This is because there is no mechanism in the industry to classify data as one service flow and another priority service flow. Therefore, all data flows through the first upstream service flow. The DOCSIS 1.1 standard provides a classification feature, activated at setup, to direct the transmission of packets at the cable modem to a particular service flow. For example, the classification could be set to put all voice packets into the second service flow while all other data packets are transmitted through the first service flow.
[0020]
According to the invention, the first service flow is divided into a number of sub-flows. Each of the subflows is assigned to a service class according to a related ranking or a priority scheme that determines the ranking for transmission standby data on the subflow from the cable modems 105 and 106 to the headend device 101. Higher priority data is transmitted before lower priority data. For example, IP phone calls need to be transmitted in near real time to avoid delays and disrupted conversations. Delays of about 300 milliseconds or more are known to be generally unbearable by the parties to the conversation. Therefore, data shown on an IP phone or the like will be assigned a higher priority service class. Conversely, text data will be assigned a lower priority service class.
[0021]
Service flow parameters (such as transmission priority) provided by the modem are set by a configuration file. Subflow parameters are generated from service flow parameters, as described below.
[0022]
In one embodiment of the present invention, the token bucket method manages upstream subflows carrying data packets, and guaranteed packets are transmitted by cable modems 105, 106 according to assigned service priorities. Each token bucket is made up of three components: emission, medium rate, and time interval. The intermediate (average) rate specifies how much data can be transmitted per unit time defined by the time interval. The emission volume (measured in bits) specifies the amount of data transmitted in a single emission to prevent network schedule conflicts from occurring. A typical token bucket algorithm using 5 Mbps token bucket emissions consists of an intermediate bit rate of 1 Mbps and a time interval of 1 second. The bucket has a maximum capacity of 5 megabits. When the bucket is full (including 5 megabit worth of tokens), no more tokens are added.
[0023]
Each token represents a permission for the cable modem to transmit a certain number of bits towards the headend device 101. To transmit the packet, the cable modem removes as many tokens from the bucket as the indication for the packet size. If there are not enough tokens in the bucket to send the packet, the packet waits until the bucket has enough tokens or the packet is simply dropped. Therefore, at any given time, the largest emission or packet that a cable modem can transmit in the network is roughly proportional to the bucket capacity.
[0024]
Further in accordance with the teachings of the present invention, the transmission of packets from cable modems 105, 106 is controlled by a number of token buckets, each token bucket being configured by a particular class of service (subflow) for the cable modem. Tokens from the first flow bucket are distributed among many subflow buckets. As described above, tokens are replenished periodically from the first bucket. Each subflow bucket is replenished from the first token bucket. In fact, there are many algorithms for refilling subflow buckets. For example, the first bucket can be controlled by allocating a defined percentage of received talk for each subflow. For example, the highest priority subflow bucket (bucket one) may receive 50% of the tokens in the first bucket over any time interval. The lowest priority subflow bucket (bucket three) receives 15% of the tokens in the first bucket over the same time interval. Finally, the middle priority bucket (bucket tow) receives 35% of the first bucket token over any time interval. Alternatively, if the first bucket contains a token, it can be placed first in bucket one (the highest priority subflow bucket) until it is full. A high priority packet guarantees a high probability of receiving the token and a high priority of transmission. Tokens received later are placed in bucket tow until bucket tow is full (assuming bucket one is full), and finally, tokens arriving if the other two buckets are full are placed in bucket three. You. As will be appreciated by those skilled in the art, various techniques are available for managing the replenishment process for either the first bucket or many subflow buckets.
[0025]
The algorithm for refilling the first bucket and many subflow buckets may be hard coded into the cable modems 105, 106, and the algorithm may be set in a configuration file. Through the vendor-specific TLV parameters, the configuration file is downloaded to the cable modems 105, 106 during the startup phase. Using a configuration file for the control mechanism allows the cable system operator to change the replenishment algorithm for multiple subflows. Also, the cable system operator can manage each of the cable modems 105 and 106 by using network management application software that runs under Simple Network Management Protocol (SNMP). SNMP allows a network manager to modify network objects in a network device, thereby changing network parameters. For example, the bucket replenishment process can be modified in real time by SNMP-based commands.
[0026]
In one implementation, each data packet is assigned to one of the subflows according to information contained in the packet header. For example, the packet header identifies IP data, LLC data, and TCP / UDP data. Based on the data type, data packets are assigned to sub-flows of appropriate priority, and the correlation between data type and sub-flow priority is set in a configuration file or hard-coded in the modem. Alternatively, to identify the data type, the modem can examine the data file for a special data signature. VoIP data has such a special signature. In another implementation, the modem can be assigned to a subflow by identifying the data type and examining the built-in cable modem software application that generates the packet. For example, a network management application generates a Simple Network Management Protocol (SNMP) packet. Typically, SNMP packets are assigned a higher priority than data packets.
[0027]
Instead of examining the header for the cable modem subflow assignment, the application software that creates the packet can assign the subflow and provide these parameters. In another example, the modem includes a VoIP software application that digitizes the telephone audible tone and packetizes it. In this example, the check for determining the priority of the packet is not necessarily required. This is because the software operating the modem can determine that the packet comes from the software application that operates the voice data. As such, these packets are generally assigned a higher priority subflow.
[0028]
FIG. 3 shows a flowchart of the software of the present invention. The modems 105 and 106 are activated in step 180, and download the setting file in step 182. In the practice of the present invention, details of the plurality of subflows and the bucket replenishment algorithm are identified in the configuration file. This information is generated in the configuration file and provided to the modem operating software which manages the replenishment process and subflow assignment. In step 186, the packet examines the service class determined with the packet in preparation for transmission. As discussed in other examples, the class of service can be determined by techniques other than header inspection. In any case, once the service class for the packet has been determined, the process moves to step 188 where the subflow (or service class) bucket constituted by the packet is examined. If enough tokens are available in the bucket, the process moves from decision step 190 to step 192 where the appropriate number of tokens are assigned to the packet. At step 194, the packet is transmitted. The process then returns to step 186 to check for the next waiting packet. If a sufficient number of tokens are not available at decision step 190, the packet cannot be forwarded and is kept in a queue. The process returns from decision process 190 to step 186 for examining the next queued packet. The packet is transmitted when the subflow token bucket composed of the delayed packet is replenished. As discussed above, the replenishment process periodically places a token in each subflow bucket, which is not described in the flowchart of FIG. In one embodiment, the replenishment process is performed as follows. Let R / T be the intermediate rate of the algorithm. R is measured in bits and T is measured in seconds. Then, a token corresponding to R bits every T seconds is added to the bucket. For example, a token equivalent to R = 1 million bits, T = 1 second, and 1 million bits per second is added to the bucket. At this time, the intermediate rate is 1 Mbps.
[0029]
As described above, the cable system operator can modify the operation of the subflow allocation process with the transmission network management control information (using the SNMP protocol) for the cable modems 105, 106. Such management information operates as an interruption to the flowchart of FIG.
[0030]
Having described the invention by way of example in the preferred embodiment, those skilled in the art will recognize that various modifications and changes can be made without departing from the scope and spirit of the invention. Changes can be made to adapt to a particular embodiment without departing from the scope of the invention. The invention is not limited only to the implementation considered the best mode disclosed, but only by the appended claims and their alternatives.
[Brief description of the drawings]
FIG. 1 is an electrical block diagram showing a typical prior art two-way cable communication system.
FIG. 2 is an electrical block diagram illustrating elements of a two-way cable communication system at a subscriber premises.
FIG. 3 is a flowchart illustrating service class assignment and methodology related to the present invention.

Claims (18)

In a network consisting of a first service flow comprising a plurality of network devices for transmitting and receiving packets,
Dividing the first service flow into a plurality of subflows;
Assigning a priority to each of the plurality of subflows;
Assigning a priority to each packet;
Transmitting to the assigned priority-based subflow;
For managing packet transmissions. The method of claim 1, wherein each priority is associated with a class of service. 3. The method of claim 2, wherein each service class is associated with a data transmission priority, and each packet is transmitted according to a service class assigned to the packet. The method of claim 1, wherein the priority is associated with a defined transmission bit rate. The method of claim 1, wherein each packet is queued before transmission, and a priority is associated with a prescribed maximum time between a packet arriving at the queue and a packet transmitted on a corresponding subflow. The method of claim 1, wherein parameters from a hierarchy of network transmission parameters are associated with each priority. The method of claim 1, wherein transmission of a packet in each subflow of a plurality of subflows requires assignment of a number of tokens to the packet, wherein the number of tokens is based on a packet length. The method of claim 1, wherein the first service flow is assigned a number of tokens, and then the token is assigned among a plurality of subflows. 9. The method of claim 8, wherein tokens are allocated among a plurality of subflows based on a number of unused tokens in the subflow. 9. The method of claim 8, wherein tokens are assigned among the plurality of subflows based on a priority associated with the subflow. A token is assigned to the first service, then the token is distributed to multiple subflows, and the highest priority service flow receives a specified percentage of the token from the first service flow, and assigns each lower priority 9. The method according to claim 8, wherein the service flows also receive a smaller percentage of assignments. The first service flow comprises a token bucket associated with the first service flow, each subflow comprises a token bucket associated with the subflow, each token bucket being defined by an emission rate, an intermediate rate, and an emission quantity being defined by each emission. 9. The method of claim 8, wherein the maximum amount of data that can be transmitted is defined, and the intermediate rate defines the maximum amount of data that can be transmitted for a defined time interval. 13. The method of claim 12, wherein the tokens of the first service flow token bucket are periodically replenished, and each subflow token bucket is replenished from the first service flow. The method of claim 1, wherein the priority of each packet is set in a packet header. The method of claim 1, wherein the priority of each packet is determined based on a software application that generates the packet. The method of claim 1, wherein the network device comprises a cable modem. A computer readable medium for implementing management of packet transmission in a network in which the computer usable medium comprises the first service flow, wherein the network comprises a plurality of network devices for packet transmission and reception. ,
Computer readable program code for configuring a computer to distribute a first service flow to a plurality of subflows;
Computer readable program code for setting a computer to assign a priority to each of the plurality of subflows;
Computer readable program code for setting the computer to assign a priority to each packet;
Computer readable program code for configuring a computer to transmit packets on a subflow associated with the assigned priority;
Products consisting of. An apparatus in each of the plurality of network devices, wherein the network device transmits and receives packets over the network comprising the first service flow;
A first module for separating a first service flow into a plurality of subflows,
A second module that assigns priority to each subflow of the plurality of subflows, a third module that assigns priority to each packet, and a fourth module that transmits packets to subflows associated with the assigned priority,
Device that manages packet transmission.

Images