JP4472674B2 - Backoff protocol optimal control method, base station, and terminal station - Google Patents

Description

  The present invention relates to a backoff protocol optimal control method, a base station, and a terminal station in a wireless communication system in which the amount of random access traffic for transmitting packets is controlled by a backoff protocol.

  As a random access method using a back-off protocol for transmission control, there is a WLAN (Wireless Local Area Network, wireless LAN). The back-off protocol is a protocol for controlling the collision caused by the simultaneous start of packet transmission from the terminal station. In a DCF (Distributed Coordination Function) mechanism that is autonomous random access provided by WLAN, a binary exponential backoff protocol using 2 as a backoff factor is adopted. In the binary exponential backoff protocol, the terminal station generates random numbers that are uniformly distributed within the range of [0, W−1] defined by the value (W) of the contention window, and at regular intervals. In this method, the generated random number value is subtracted, transmission is performed when the random number value becomes 0, and the contention window value is doubled according to the number of retransmissions. However, since the initial value of the contention window is fixed, the throughput and delay characteristics deteriorate as the number of terminals increases.

Therefore, many mechanisms have been proposed for controlling the initial value of the contention window to the optimum value according to the traffic situation. In the methods described in Non-Patent Document 1 and Non-Patent Document 2, each terminal station autonomously estimates the number of terminals (active terminals) with transmission waiting packets, and the initial contention window is proportional to the number. Apply the value.
In the method described in Non-Patent Document 3, each terminal station autonomously monitors the channel state, and the ratio between the number of slots in which a collision has occurred and the number of empty slots that have not been transmitted from any terminal station Are used as parameters, the initial value of the contention window is increased if the value is larger than the optimum value, and the initial value of the contention window is decreased if the value is smaller.
G. Bianchi, L. Fratta, M. Oliveri, “Performance Evaluation and Enhancement of the CSMA / CA MAC Protocol for 802.11 Wireless LANs) ", IEEE PIMRC 1996, pp. 392-396 Y. Chen, QA Zeng, D. Agrawal, “Performance Analysis and Enhancement for IEEE 802.11 MAC Protocol”, ICT 2003, pp. 860-867 Y. Peng, H. Wu, S. Cheng, K. Long, “A New Self-Adapt DCF Algorithm”, IEEE GLOBECOM 2002, pp. 197 87-91

  As described above, in a system in which a large number of terminals randomly access, the channel usage efficiency (throughput) is greatly reduced unless transmission control is properly performed by the back-off protocol. A control method has been proposed.

  In order to perform transmission control appropriately using the back-off protocol, it is necessary to accurately grasp the traffic situation. However, in many conventional techniques (Non-Patent Documents 2 and 3), the number of active terminals is estimated by grasping the results of access in each random access slot in three states of success / collision / vacancy. However, it is difficult to accurately determine the collision / vacancy state.

  Even if the traffic situation can be accurately grasped, in many conventional technologies (Non-patent Documents 1 and 2), if a plurality of terminal stations transmit packets at the same time, the packets are always destroyed. The initial value of the optimal contention window is calculated. Alternatively, in the method of Non-Patent Document 3, the optimum value of the ratio of the collision slot and the empty slot is derived on the premise that if a plurality of terminal stations transmit packets at the same time, the packets are always collapsed, and the optimum value is calculated. Use to control. However, in the actual case, even if simultaneous transmission of packets occurs, if the intensity of other packets other than a certain packet is relatively weak enough, a capture effect that is correctly demodulated occurs. For this reason, in a situation where many capture effects occur, the calculated initial value of the contention window is not necessarily the optimum value.

  The present invention has been made in view of such circumstances, and its purpose is a parameter of the back-off protocol in a system in which the amount of random access traffic for transmitting packets is controlled by the back-off protocol. An object of the present invention is to provide a back-off protocol optimal control method, a base station, and a terminal station that can adaptively control an initial contention window and a back-off factor according to traffic conditions.

  In order to solve the above-mentioned problem, the invention according to claim 1 is a back-off protocol optimal control method in a wireless communication system comprising a plurality of communication devices and controlling the amount of random access traffic by a back-off protocol, A communication device that receives random access from another communication device estimates the supply traffic amount and transmission failure probability from the other communication device, and estimates the current supply traffic amount and transmission failure probability, and the supply traffic estimated in the past The target supply traffic amount is calculated from the amount and the transmission failure probability, and if the calculated target supply traffic amount is larger than the current supply traffic amount, the initial value of the current contention window is decreased, and the calculated target supply is calculated. If the traffic volume value is less than the current supply traffic volume, the current content The initial value of the contention window is increased, the initial value of the contention window is notified to the other communication device, and the other communication device that performs random access uses the initial value of the contention window received from the communication device to set the backoff time. This is a back-off protocol optimal control method characterized by determining.

  The invention according to claim 2 is the backoff protocol optimum control method according to claim 1, wherein the other communication device performing random access embeds information indicating the number of retransmissions in a random access packet. The communication device that transmits and receives random access from another communication device uses the information on the number of retransmissions in the random access packet received from the other communication device, and calculates the throughput for each number of retransmissions to supply traffic. And estimating a transmission failure probability.

  The invention described in claim 3 is the backoff protocol optimal control method according to claim 1, wherein the communication device receiving random access from another communication device is configured to estimate the supplied traffic amount and the transmission failure probability. Is recorded in the recording unit, and the supply traffic amount and transmission failure probability data estimated in the past that satisfy a predetermined condition for calculating an appropriate supply traffic amount are read from the recording unit, and the target supply The traffic volume is calculated.

  The invention according to claim 4 is the backoff protocol optimum control method according to claim 1, wherein the communication device receiving random access from another communication device cannot calculate a target supply traffic amount, In addition, when the target supply traffic volume has been calculated so far, the initial value of the contention window is changed by comparing the target supply traffic volume calculated in the past with the current supply traffic volume. And

  Further, the invention according to claim 5 is the backoff protocol optimum control method according to claim 1, wherein the communication device receiving random access from another communication device can calculate a target supply traffic amount. If the target supply traffic volume has not been calculated up to now, the initial value of the current contention window is changed and notified to the other communication device, and after a predetermined time, the target supply traffic is again returned. The amount of traffic is calculated, the initial value of the contention window is changed, and another communication device is notified.

  The invention according to claim 6 is the backoff protocol optimal control method according to any one of claims 1 to 5, wherein the backoff factor value is set instead of the initial value of the contention window. It is characterized by changing.

  The invention according to claim 7 is the base station in a radio communication system that includes a base station and a plurality of terminal stations, and controls the amount of random access traffic from the terminal stations by a back-off protocol, Estimate the supply traffic amount and transmission failure probability from the terminal station, and calculate the target supply traffic amount from the estimated current supply traffic amount and transmission failure probability, and the previously estimated supply traffic amount and transmission failure probability. If the calculated target supply traffic amount is larger than the current supply traffic amount, the initial value of the current contention window is reduced. If the calculated target supply traffic amount is smaller than the current supply traffic amount, the current A control unit for increasing the initial value of the tension window, and a content changed by the control unit. A control signal generating unit for notifying ® emission window initial value to a terminal station, a base station, characterized in that it comprises a.

  The invention according to claim 8 is the base station according to claim 7, wherein the control unit changes a value of a back-off factor instead of an initial value of a contention window, and the control signal The generation unit notifies the terminal station of the value of the back-off factor changed by the control unit.

  The invention of claim 9 is the terminal station in a radio communication system comprising a base station and a plurality of terminal stations, and controlling the amount of random access traffic from the terminal station by a back-off protocol, wherein the base station A retransmission number management unit that counts the number of retransmissions of a random access packet transmitted to the packet generation unit that embeds and transmits information indicating the number of retransmissions counted by the retransmission number management unit in a random access packet; and And a back-off time management unit that determines a back-off time according to the received contention window initial value.

  The invention according to claim 10 is the terminal station according to claim 9, wherein the back-off management unit receives and receives a back-off factor value instead of the initial contention window value. The back-off time is determined by the value of the back-off factor.

In the optimal control method of the back-off protocol of the present invention, the traffic situation is monitored with two states of success / failure, and the initial value of the optimum contention window is calculated in consideration of the occurrence of the capture effect. .
The initial value of the contention window that maximizes the throughput is proportional to the number of active terminals. However, the proportionality factor varies depending on the probability of the capture effect. The occurrence probability of the capture effect depends on the modulation scheme and the distance between the base station and the terminal station. FIG. 1 shows an example of the relationship between the number of active terminals N and the optimum contention window initial value W when there is no capture effect and when there is no capture effect. As shown in the figure, the initial value of the calculated contention window is about three times the optimum value when the proportional coefficient when there is no capture effect in the situation where the capture effect is present.

  Therefore, in the present invention, the supply traffic amount G and the transmission failure probability Pf are estimated, and the optimum content window initial value is derived using these. The following relationship exists between the supply traffic amount G and the transmission failure probability Pf, and the throughput S.

  S = G (1-Pf) (Formula 1)

（以下、便宜上「目標供給トラヒック量Ｇ^・」と記載）は、時刻ｔにおける供給トラヒックの推定値Ｇ_ｔと、送信失敗確率の推定値Ｐｆ_ｔに基づき以下の式（２）ように算出され、この算出された目標供給トラヒック量Ｇ^・は、スループットが最大となる時の供給トラヒック量ＧをｏｐｔＧとすると、Ｇ_ｔとｏｐｔＧとの間に位置する（Ｇ_ｔ＜Ｇ^・＜ｏｐｔＧ、もしくは、ｏｐｔＧ＜Ｇ^・＜Ｇ_ｔ）。そのため、供給トラヒック量Ｇを目標供給トラヒック量Ｇ^・に誘導することで、ｏｐｔＧに近づけることができる（請求項１）。

(Hereinafter, for convenience as a "target supply amount of traffic G ^·") is the estimated value G _t of feed traffic at time t, the estimated value Pf _t on basis of the following equation of a transmission failure probability (2) is calculated as, The calculated target supply traffic amount G ^· is located between G _t and optG (G _t <G ^· <optG or optG, where optG is the supply traffic amount G when the throughput is maximum). <G ^· <G _t ). Therefore, the supply traffic amount G can be brought close to optG by guiding the supply traffic amount G to the target supply traffic amount G ^· (Claim 1).

Here, assuming that the probability of failure due to collision with k interference waves is P _k , and the number of terminals is N, the transmission failure probability Pf is expressed as (Equation 3) below. Therefore, the transmission failure probability Pf can be represented by the supply traffic amount G and the probability P _{k of} failure due to collision with k interference waves. In other words, the relationship between the supply traffic amount G and the transmission failure probability Pf is determined only by the probability P _k that collides with the k interference waves and fails, and is not affected by the occurrence traffic amount.

Since the modulation scheme does not change frequently, assuming that _Pk does not change significantly, determining the supply traffic amount G from the past history of the supply traffic amount G and the transmission failure probability Pf, the transmission failure probability Pf and Throughput S can be estimated. Therefore, the supply traffic amount G and the transmission failure probability Pf estimated so far can be recorded, and the history information can be used to calculate dPf / dG in (Equation 2) (Claim 3). Even if the target supply traffic amount G ^· cannot be calculated, the previously calculated target supply traffic amount G ^· can be utilized (claim 4). Further, if the information of the estimation history of the supply traffic amount G and the transmission failure probability Pf is not prepared, the target traffic amount G and the transmission failure probability Pf are changed by changing the parameters of the backoff protocol. The supply traffic amount G ^· can be calculated (claim 5).

Further, in order to accurately grasp the traffic situation, it is difficult to discriminate, and the collision and the empty slot that cause the traffic situation grasp error are not discriminated, and the traffic situation is estimated using the information of the success and the unsuccessful slot. That is, it is possible to measure all the throughputs S _r for each number of retransmissions (r) in addition to the overall throughput S, and to grasp the traffic situation more accurately even in a situation where the capture effect is occurring (claim 2). Figure 2 is a transmission failure probability Pf estimated using the supplied traffic G and overall throughput S _r, a transmission failure probability Pf estimated using only the supply traffic G and the number of retransmissions 0 th throughput S ₀ An example of comparison is shown. Here, the fluctuation of received power from each terminal station is ± 5 dB, and the SIR (Signal to Interference Ratio) -PER (Packet Error Rate) characteristic is PER = 1 for SIR <1 dB, PER = 0 for 1 dB <SIR <2 dB. 75, 2 dB <SIR <3 dB, PER = 0.35, 3 dB <SIR <4 dB, PER = 0.08, 4 dB <SIR <5 dB, PER = 0.000, 5 dB <SIR <6 dB, PER = 0.006 did. Pf can be estimated more accurately by using all the throughputs for each number of retransmissions.

Finally, there are two backoff protocol parameters: a contention window and a backoff factor. FIG. 3A shows the relationship between the initial value W of the contention window and the supplied traffic amount G when the number of terminals is 10,000, the generated traffic volume is 1.0, and the back-off factor is fixed to 2. FIG. 6B shows the relationship between the backoff factor value Bf and the supply traffic amount G when the initial contention window value W is fixed at 100. FIG. Since the supply traffic amount G is a decreasing function of any parameter of the initial value W of the contention window and the backoff factor value Bf, the supply traffic amount G can be derived using either parameter. That is, if the calculated G ^· is larger than G _t , the initial value W or the back-off factor value Bf of the current contention window is reduced, and if G ^· is smaller than G _t , the initial value of the current contention window W or the back-off factor value Bf is increased.

  According to the present invention, since a large number of terminals perform random access, the channel utilization efficiency (throughput) is greatly reduced and the delay time is greatly increased unless proper transmission control is performed by the back-off protocol. Even when the capture effect occurs, it is possible to appropriately control the parameters of the back-off protocol according to the traffic situation, and to suppress the deterioration of the throughput characteristic and the delay characteristic.

[First embodiment]
A first embodiment of the present invention will be described with reference to FIGS. 4 to 7, 12 and 13. FIG. 4 is a block diagram showing a network configuration of the wireless communication system of the present embodiment. FIG. 5 shows a basic sequence between the base station and the terminal station in the present embodiment. FIG. 6 is a diagram illustrating an operation flow in the base station. FIG. 7 is a diagram showing an operation flow in the terminal station. FIG. 12 is a functional block diagram of the base station, and FIG. 13 is a functional block diagram of the terminal station.
In the following description, a communication device that receives random access from another communication device is described as a base station, and a communication device that performs random access is described as a terminal station.

  As shown in FIG. 4, a plurality of terminal stations 20 (here, terminal stations # 1 to # 5) are wirelessly connected to one base station 10 to ensure bidirectional communication. On the road, the amount of random access traffic from the terminal station 20 to the base station 10 is controlled by a back-off protocol. The reception areas A0 to A5, which are each circle centered on the base station 10 and the terminal stations # 1 to # 5, are the terminal stations for the base station 10 and the terminal stations # 1 to # 5 that are the centers of the circles. A reception area for a transmission packet from 20 is shown. In the figure, the terminal station # 1 can receive the transmission packet of the adjacent terminal station # 2, but many other terminal stations # 3 to # 3 existing in the reception area A0 of the base station 10 (AP: access point). # 5 transmission packet cannot be received. In this case, even if the terminal stations # 3 to # 5 randomly access the base station 10 (AP), the terminal station # 1 cannot receive it. Therefore, the terminal station # 1 cannot accurately grasp the traffic situation of random access to the base station 10.

  Therefore, the base station 10 grasps the traffic situation, calculates the initial value W of the optimum contention window according to the traffic situation, and broadcasts the value to the terminal station 20 periodically. FIG. 5 shows a sequence between the base station 10 and the terminal station 20. In the terminal station 20 in which traffic newly occurs, the back-off time is determined by the control signal including the initial value W of the contention window received thereafter. The back-off time is subtracted for each slot, and random access is performed when it becomes zero. The success or failure of random access in the terminal station 20 is determined by the presence or absence of ACK included in the control signal. When ACK is not included, the back-off time is determined based on the latest contention window initial value, the number of retransmissions, and the value of the back-off factor.

  In the figure, the base station 10 broadcasts to the terminal station 20 a control signal in which the initial value W of the contention window and the ACK for random access from the terminal station 20 are set (step S110). Assume that after a random access is performed from one terminal station 20 to the base station 10 (step S120), a random access is performed from a plurality of other terminal stations 20 to the base station 10 and a collision occurs (step S130). The base station 10 calculates the optimal initial value W of the contention window according to the traffic situation (step S140), and transmits the control signal in which the calculated value W and the ACK for random access are set to the terminal station 20. (Step S150). In each terminal station 20 that has recognized that the random access has not reached the base station 10 by the control signal, the back-off time is determined by the initial value W of the contention window included in the control signal received in step S150 and updated. (Step S160). Furthermore, in the next cycle, the base station 10 calculates the initial value W of the optimum contention window according to the new traffic situation (step S170), and calculates the calculated value W and the ACK for random access. The set control signal is broadcast to the terminal station 20 (step S180). Each terminal station 20 subtracts the back-off time determined in step S160 for each slot, and performs random access again when it reaches zero (step S190).

FIG. 12 shows a functional block diagram of the base station 10. The base station 10 includes a reception unit 11, a transmission unit 12, a transfer unit 13, a W (Bf) control unit 14, a control signal generation unit 15, a reception unit 16, and a transmission unit 17. The receiving unit 11 receives a signal from the terminal station 20 by radio, and the transmitting unit 12 transmits the signal to the terminal station 20 by radio. The receiving unit 16 receives a signal from another node (not shown) by wire or the like, and the transmitting unit 17 transmits a signal to another node (not shown) by wire or the like. The transfer unit 13 transfers the received packet and the transmission packet to an appropriate function unit.
The W (Bf) control unit 14 further includes an S measurement unit 141, a W (Bf) calculation unit 142, and a recording unit 143. Based on the number of random access packets received from the terminal station 20 by the receiving unit 11, the S measuring unit 141 measures the throughput S, and the W (Bf) calculating unit 142 determines the optimum contention window initial value W according to the traffic situation. Alternatively, a back-off factor value Bf is calculated. The optimal contention window initial value W or backoff factor value Bf calculated by the W (Bf) calculation unit 142 is recorded by the recording unit 143 and used in the next calculation process by the W (Bf) calculation unit 142. . The contention window initial value W or backoff factor value Bf calculated by the W (Bf) calculation unit 142 and the success or failure of random access are implemented in the control signal in the control signal generation unit 15, and are transmitted via the transmission unit 12. Broadcast to the terminal station 20.

  FIG. 13 shows a functional block diagram of the terminal station 20. The terminal station 20 includes a reception unit 21, a retransmission number management unit 22, a backoff time management unit 23, a packet generation unit 24, and a transmission unit 25. The receiving unit 21 receives signals wirelessly, and the transmitting unit 25 transmits signals wirelessly. When the packet generation unit 24 generates a packet, the retransmission number management unit 22 clears the retransmission counter held therein to zero. Thereafter, the reception unit 21 receives a control signal including the contention window initial value W from the base station 10, and the back-off time management unit 23 determines the back-off time based on the information. The back-off time management unit 23 subtracts the back-off time for each slot time, and sends a packet transmission instruction to the packet generation unit 24 when it becomes zero. The packet generation unit 24 embeds the value of the retransmission counter managed by the retransmission number management unit 22 and transmits the packet. After the transmission, the control signal including the ACK is received by the retransmission number management unit 22 via the reception unit 21 and it is confirmed whether or not the transmitted packet has reached the base station 10. If it is determined that the number of retransmissions is not reached, the retransmission number management unit 22 increments the retransmission counter by 1 and instructs the backoff time management unit 23 to determine the backoff time again.

FIG. 6 shows a flow of calculation operation of the contention window initial value W in the base station 10 of the present embodiment. The W (Bf) calculation unit 142 estimates the supply traffic amount G and the transmission failure probability Pf for each update period of the contention window initial value (the calculation period of the throughput S) (step S210). ) To calculate the target supply traffic amount G ^· (step S220). dPf / dG is the time (t-1) supplied traffic was estimated _{G t-1} and transmission failure probability _{Pf t-1} the immediately preceding estimated in step S210, _{G t} and supply traffic of the current time t approximated with a transmission failure probability Pf _t the rate of change ΔPf / ΔG from the difference of _{= (Pf t -Pf t-1} ) / (G t -G t-1).

Supply traffic G is represented by a decreasing function of the contention window initial value (dG / dW <0) for, W (Bf) calculation unit ^142, G ·> the initial value of the contention window if _{G t} W a small predetermined value W_step than the current value (step S230: Yes, step ^S250), increased by a predetermined value W_step than the current value if G · _{<G t} (step S230: No, step S240) . Or, <order (0, W (Bf) calculation unit ^142, G · supply traffic G is represented by a decreasing function of back-off factors dG / dBf)> The back-off factors of value Bf if _{G t} It is smaller than the current value by a predetermined value, and if G ^· < _Gt , it is larger than the current value by a predetermined value.

  If the initial value W of the contention window newly calculated in step S250 is smaller than the predetermined lower limit of the initial value W of the contention window (step S260: Yes), the W (Bf) calculation unit 142 The initial value W of the new contention window is set as the lower limit value (step S270).

  FIG. 7 shows an operation flow in the terminal station 20 of the present embodiment from when a packet occurs until the random access is completed. If a new packet is generated in the packet generation unit 24 (step S310), the retransmission number management unit 22 of the terminal station 20 first clears the retransmission counter r held therein to zero (step S320). Thereafter, when the back-off time management unit 23 receives a control signal including the initial value W of the contention window from the base station 10 via the reception unit 21 (step S330), within the range of [0, W−1]. A back-off time is determined (step S340). The back-off time management unit 23 subtracts the back-off time for each slot time (step S350, step S360: No), and outputs a transmission instruction to the packet generation unit 24 when it becomes zero (step S360: Yes), the packet generator 24 transmits a random access packet (step S370).

After the transmission, the retransmission number management unit 22 receives a control signal including ACK from the base station 10 via the reception unit 21, and whether its own ID or PE (partial echo) is set in the received control signal. If no search is found, it is determined that the random access packet has not reached the base station 10 (step S380: Yes), and the retransmission counter r is incremented by 1 (step S390). If the retransmission counter r has reached the upper limit value (step S400: Yes), the back-off time management unit 23 discards the random access packet (step S410). On the other hand, if the retransmission counter r has not reached the upper limit value (step S400: No), the back-off time management unit 23 is based on the initial value W of the latest contention window, the back-off factor value Bf, and the retransmission counter r. , [0, Bf ^r W−1], a back-off time is determined, and the process from step S330 is repeated to perform back-off again. Note that when the packet is discarded (step S410) or when the success of random access is confirmed by receiving an ACK (step S380: No), the packet generation unit 24 is as long as there is a transmission packet in the queue (step S380). S420: No), the processing from step S320 is executed, and random access is repeated again.

[Second Embodiment]
The second embodiment of the present invention will be described focusing on the differences from the first embodiment described above with reference to FIG.

  FIG. 8 shows an operation flow in the terminal station 20 of the present embodiment from when a packet occurs until the random access is completed. When a new packet is generated in the packet generator 24 (step S510), the retransmission count manager 22 of the terminal station 20 first clears the retransmission counter r held therein to zero (step S520). Thereafter, when the back-off time management unit 23 receives a control signal including the initial value W of the contention window from the base station 10 via the reception unit 21 (step S530), the back-off time management unit 23 falls within the range of [0, W−1]. A back-off time is determined (step S540). The back-off time management unit 23 subtracts the back-off time for each slot time (step S550, step S560: No), and outputs a transmission instruction to the packet generation unit 24 when it becomes zero (step S560: Yes), the packet generator 24 embeds the value of the retransmission counter r in the random access packet and transmits it (steps S570 and S580).

After the transmission, the retransmission number management unit 22 receives a control signal including ACK via the reception unit 21 and searches for whether its own ID or PE is set in the received control signal. It is determined that the access packet has not reached the base station 10 (step S590: Yes), and the retransmission counter r is incremented by 1 (step S600). If the retransmission counter r has reached the upper limit value (step S610: Yes), the backoff time management unit 23 discards the random access packet (step S620). On the other hand, if the retransmission counter r has not reached the upper limit (step S610: No), the back-off time management unit 23 is based on the initial value W of the latest contention window, the back-off factor value Bf, and the retransmission counter r. , [0, Bf ^r W−1] is determined within the range, and the processing from step S530 is repeated to perform backoff again. Note that if the packet is discarded (step S620), or if the success of random access is confirmed by receiving an ACK (step S590: No), the packet generation unit 24 is as long as there is a transmission packet in the queue ( Step S630: No), the processing from step S520 is executed, and random access is repeated again.

On the other hand, the S measuring unit 141 of the base station 10 refers to the value of the retransmission counter r included in the random access packet received from the terminal station 20 and measures the throughput S _r for each retransmission count r. Here, if traffic occurs in a terminal station k, which is a certain terminal station 20, and the probability of transitioning to the first transmission state by performing backoff is B _k and the transmission failure probability is Pf _k , the number of transmissions r The throughput S _{r of} (r = 0 to R−1) and the overall throughput S can be expressed as the following (Formula 5).

Therefore, the supply traffic amount G can be estimated using the measured _Sr as in the following (Equation 6).

  Using the estimated G and the measured S, the transmission failure probability Pf can be estimated as in the following (Equation 7).

  In step S210 of FIG. 6 of the first embodiment, the base station 10 estimates the supply traffic amount G and the transmission failure probability Pf as described above, and uses the estimated supply traffic amount G and the transmission failure probability Pf as described above. The same processing as that after step S220 in FIG. 6 is executed.

[Third embodiment]
A third embodiment of the present invention will be described with a focus on differences from the first embodiment described above with reference to FIG.

FIG. 9 is a block diagram showing an operation flow in the base station 10 of the present embodiment.
Here, if the probability of failure due to collision with k interference waves is P _k , and the number of terminals is N, the transmission failure probability Pf is expressed as in the following (Equation 8). Therefore, the transmission failure probability Pf can be represented by the supply traffic amount G and the probability _{Pk of} colliding with k interference waves and failing. In other words, the relationship between the supply traffic amount G and the transmission failure probability Pf is determined only by the probability P _{k of the} collision failure with the k interference waves and is not affected by the generated traffic amount.

On the other hand, as in the first embodiment ^, in calculating the target supply traffic amount G ^· , dPf / dG is obtained by the rate of change ΔPf / ΔG from the adjacent values of the supply traffic amount G and the transmission failure probability Pf. . Since the estimated supply traffic amount G and the transmission failure probability Pf include an estimation error, ΔPf / ΔG calculated when the value of the supply traffic amount G and the transmission failure probability Pf is too close in calculating ΔPf / ΔG is a large error. Will be included.

In FIG. 9, the W (Bf) calculation unit 142 estimates the supply traffic amount G and the transmission failure probability Pf in the update period of the contention window initial value as in the first embodiment, and the estimated supply traffic amount G and transmission The failure probability Pf is recorded in the recording unit 143 (step S710). The W (Bf) calculation unit 142 supplies the traffic that has been separated by more than a specified value that is sufficiently larger than the estimation error with respect to the supply traffic amount G _t and the transmission failure probability Pf _{t at} the current time t estimated in step S710. The past data of the amount G and the transmission failure probability Pf is extracted from the recording unit 143 as reference data (step S720), and these are used as the supply traffic amount G _t-1 and the transmission failure probability Pf _{t-1 to} be used in the first embodiment. The target supply traffic amount G ^· is calculated by the same procedure as (step S730). Subsequent steps S740 to S780 execute the same processing as steps S230 to S270 of FIG. 6 of the first embodiment.

[Fourth embodiment]
A fourth embodiment of the present invention will be described focusing on differences from the third embodiment described above with reference to FIG.

FIG. 10 is a diagram illustrating an operation flow in the base station 10 of the present embodiment.
If the probability P _{k of} failure due to collision with k interference waves is substantially constant, the relationship between the supply traffic amount G and the transmission failure probability Pf does not change, so the target supply traffic amount G ^· calculated so far is used. However, the initial value of the contention window can be controlled sufficiently. Therefore, the W (Bf) calculation unit 142 calculates before the data of the supply traffic amount G and the transmission failure probability Pf that match the extraction conditions of the reference data does not exist in the database in the procedure in the third embodiment. The target supply traffic amount G ^· is read and used.

That is, the W (Bf) calculation unit 142 estimates the supply traffic amount G and the transmission failure probability Pf in the update period of the contention window initial value by the same processing as step S710 of FIG. 9 in the third embodiment, and estimates The supplied traffic amount G and the transmission failure probability Pf are recorded in the recording unit 143 (step S810). Then, the W (Bf) calculation unit 142 is separated from the supply traffic amount G _{t at} the current time _t and the transmission failure probability Pf _t estimated in step S810 by a predetermined value that is sufficiently larger than the estimation error. It is determined whether past data of the supply traffic amount G and the transmission failure probability Pf is registered in the recording unit 143 (step S820). If it is not registered (step S820: No), the target supply traffic amount G ^· calculated before that is read out from the recording unit 143 and used, and if it exists (step S820: Yes), the past supply traffic amount G and the transmission failure probability Pf are read out, and the target supply traffic amount G ^· is calculated and updated by the same processing as step S730 in FIG. 9 in the third embodiment (step S830). Subsequent steps S840 to S880 execute processing similar to steps S740 to S780 of FIG. 9 of the third embodiment.

[Fifth embodiment]
A fifth embodiment of the present invention will be described focusing on differences from the fourth embodiment described above with reference to FIG.

FIG. 11 is a diagram illustrating an operation flow in the base station 10 of the present embodiment.
As in the fourth embodiment, the base station 10 calculates dPf / dG by the rate of change ΔPf / ΔG from the adjacent value of the supply traffic amount G and the transmission failure probability Pf in calculating the target supply traffic amount G ^·. And As described above, since the estimated supply traffic amount G and the transmission failure probability Pf include an estimation error, ΔPf calculated when the supply traffic amount G and the transmission failure probability Pf are too close in calculating ΔPf / ΔG. / ΔG includes a large error.

Therefore, in the present embodiment, the initial value of the contention window is changed, and the estimated supply traffic amount G and the transmission failure probability Pf are changed by a predetermined value or more based on the estimation error of the supply traffic amount G and the transmission failure probability Pf. Thus, the target supply traffic amount G ^· can be calculated. However, if the amount of generated traffic is small, the supply traffic amount G and the transmission failure probability Pf hardly change even if the initial value of the contention window is changed. Therefore, an upper limit value of the initial value of the contention window is set. Try not to make it bigger.

That is, the W (Bf) calculation unit 142 of the base station 10 performs the contention window initial value update cycle (traffic S calculation cycle) by the same processing as steps S810 and S820 of FIG. 10 in the fourth embodiment. The supply traffic amount G and the transmission failure probability Pf are estimated, the estimated supply traffic amount G and the transmission failure probability Pf are recorded in the recording unit 143 (step S910), and the reference data, that is, the current time calculated in step S910 The past data of the supply traffic amount G and the transmission failure probability Pf that are separated from the supply traffic amount G _t and the transmission failure probability Pf _t by a predetermined value that is sufficiently larger than the estimation error is registered in the recording unit 143. It is determined whether or not (step S920). When the reference data is registered (step S920: Yes), the same processing as steps S830 to S880 in FIG. 10 of the fourth embodiment is performed (steps S930 to S980).

On the other hand, when the reference data in the recording unit 143 is not present (step S820: No), earlier on whether to calculate the target supply amount of traffic G ^·, i.e. the target supply amount of traffic G ^· read from the recording unit 143 is 0 Whether or not (step S990). When the read target supply traffic amount G ^· is 0 (step S990: No), the processing after step S940 is executed. When the target supply traffic amount G ^· is 0 (step S990: Yes), a flag that is turned on when the initial value W of the contention window reaches the upper limit is read from the recording unit 143. If the read flag is OFF (step S1000: Yes), the contention window initial value W is increased by a predetermined value W_step from the current value (step S1010). When the initial value W of the contention window newly calculated in step S1010 is larger than the upper limit of the initial value W of the predetermined contention window (step S1020: Yes), the W (Bf) calculation unit 142 The initial value W of the contention window is set to the upper limit value, and the flag is set to ON (step S1030).

  Alternatively, if the read flag is not OFF (step S1000: No), the initial value W of the contention window is made smaller by a predetermined value W_step than the current value (step S1040). When the initial value W of the contention window newly calculated in step S1040 is smaller than the predetermined lower limit of the initial value W of the contention window (step S1050: Yes), the W (Bf) calculation unit 142 generates a new content window. The initial value W of the contention window is set to the lower limit value, and the flag is set to OFF (step S1060).

As described above, by changing the initial value W of the contention window, the supply traffic amount G and the transmission failure probability Pf estimated in step S910 are changed in the next cycle.
When the value of the back-off factor value Bf is changed, the initial value W of the contention window may be replaced with the back-off factor value Bf in the above processing.

  Although the embodiment of the present invention has been described above, the W (Bf) control unit 14 and the control signal generation unit 15 of the base station 10, the retransmission number management unit 22, the backoff time management unit 23 of the terminal station 20, The packet generator 24 may be realized by dedicated hardware (for example, wired logic) and is configured by a memory and a CPU (central processing unit), and loads a program from the memory. The function may be realized by executing the function. Further, the W (Bf) control unit 14 and the control signal generation unit 15 of the base station 10 and the retransmission number management unit 22, the backoff time management unit 23, and the packet generation unit 24 of the terminal station 20 are realized. Necessary processing may be performed by recording the program on a computer-readable recording medium, causing the computer system to read and execute the program recorded on the recording medium. The “computer system” here includes an OS and hardware such as peripheral devices.

Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a portable medium such as a flexible disk, a magneto-optical disk, a ROM, and a CD-ROM, and a storage device such as a hard disk built in the computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

The relationship figure of the number of active terminals and the optimal contention window initial value. The figure shown about the estimation error of a transmission failure probability. The figure which shows the relationship between the initial value of a contention window, a back-off factor, and the amount of supply traffic. 1 is a block diagram of a network configuration according to a first embodiment. The basic sequence diagram of the base station and terminal station of 1st embodiment. The operation | movement flowchart in the base station of 1st embodiment. The operation | movement flowchart in the terminal station of 1st embodiment. The operation | movement flowchart in the terminal station of 2nd embodiment. The operation | movement flowchart in the base station of 3rd embodiment. The operation | movement flowchart in the base station of 4th embodiment. The operation | movement flowchart in the base station of 5th embodiment. The functional block diagram of a base station. The functional block diagram of a terminal station.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 ... Base station 11, 16, 21 ... Reception part 12, 17, 25 ... Transmission part 13 ... Transfer part 14 ... W (Bf) control part 141 ... S measurement part 142 ... W (Bf) calculation part 143 ... Recording part 15 ... Control signal generator 20 ... Terminal station 22 ... Retransmission count manager 23 ... Backoff time manager 24 ... Packet generator

Claims (10)

A back-off protocol optimal control method in a wireless communication system comprising a plurality of communication devices and controlling a random access traffic amount by a back-off protocol,
A communication device that receives random access from another communication device
Estimate the supply traffic amount and transmission failure probability from the other communication devices, and target supply traffic amount from the estimated current supply traffic amount and transmission failure probability, and the previously estimated supply traffic amount and transmission failure probability If the calculated target supply traffic volume value is larger than the current supply traffic volume, the initial value of the current contention window is decreased, and if the calculated target supply traffic volume value is smaller than the current supply traffic volume, Increase the current contention window initial value, notify the other contention device of the contention window initial value,
The other communication device that performs random access,
A back-off time is determined by an initial contention window value received from the communication device;
Backoff protocol optimal control method characterized by that. The other communication device that performs random access is:
Embed information indicating the number of retransmissions in a random access packet and send it,
The communication device receiving random access from another communication device is:
Using the information on the number of retransmissions in the random access packet received from the other communication device, and estimating the supply traffic amount and the transmission failure probability by calculating the throughput for each number of retransmissions,
The back-off protocol optimal control method according to claim 1. The communication device receiving random access from another communication device is:
The combination of the estimated supply traffic amount and the transmission failure probability is recorded in the recording unit, and the previously estimated supply traffic amount and transmission failure probability data satisfying a predetermined condition for calculating an appropriate supply traffic amount are recorded. To calculate the target supply traffic volume,
The back-off protocol optimal control method according to claim 1. The communication device receiving random access from another communication device is:
If the target supply traffic volume cannot be calculated and the target supply traffic volume has been calculated up to now, the target supply traffic volume calculated in the past is compared with the current supply traffic volume. Change the initial value,
The back-off protocol optimal control method according to claim 1. The communication device receiving random access from another communication device is:
If the target supply traffic volume cannot be calculated and the target supply traffic volume has not been calculated so far, the initial value of the current contention window is changed and notified to the other communication device. Then, after a predetermined time, the target supply traffic amount is calculated again to change the initial value of the contention window and notify other communication devices.
The back-off protocol optimal control method according to claim 1.   6. The back-off protocol optimal control method according to claim 1, wherein the back-off factor value is changed instead of the initial value of the contention window. The base station in a wireless communication system, which comprises a base station and a plurality of terminal stations, and controls the amount of random access traffic from the terminal station by a back-off protocol,
Estimate the supply traffic amount and transmission failure probability from the terminal station, and calculate the target supply traffic amount from the estimated current supply traffic amount and transmission failure probability, and the previously estimated supply traffic amount and transmission failure probability If the calculated target supply traffic volume value is larger than the current supply traffic volume, the initial value of the current contention window is decreased, and if the calculated target supply traffic volume value is smaller than the current supply traffic volume, the current A control unit for increasing the initial contention window;
A control signal generation unit for notifying the terminal station of the contention window initial value changed by the control unit;
A base station comprising: The control unit changes the value of the backoff factor instead of the initial value of the contention window,
The control signal generation unit notifies the terminal station of the value of the back-off factor changed by the control unit;
The base station according to claim 7. The terminal station in a wireless communication system, which comprises a base station and a plurality of terminal stations, and controls the amount of random access traffic from the terminal station by a back-off protocol,
A retransmission number management unit that counts the number of retransmissions of a random access packet transmitted to the base station;
A packet generation unit that embeds and transmits information indicating the number of retransmissions counted by the retransmission number management unit in a random access packet;
A back-off time management unit that determines a back-off time according to a contention window initial value received from the base station;
A terminal station comprising: The back-off management unit receives a back-off factor value instead of the contention window initial value, and determines a back-off time according to the received back-off factor value.
The terminal station according to claim 9.

Images