TWI376123B - Fixed window backoff method capable of qos guarantee for wireless lans - Google Patents

Description

1376123 VI. Description of the Invention: [Technical Field] The present invention relates to a medium access control MAC method, and more particularly to a fixed window retreat method for service quality assurance in a day-to-day wireless area network It can be applied to wireless local area network to achieve both quality of service (QoS) and high channel utilization (channe丨Utilization). [Prior Art] A wireless local area network refers to a regional network that does not use physical lines. At present, most wireless local area network technologies use radio waves (rf) as transmission media to form a communication network. The basic service set of a wireless local area network usually includes a wireless access point (AP) and a limited number of mobile stations (such as wireless network cards, pDA, pen 31 computers, etc.) Wait). The wireless LAN system uses wireless channels to connect the various members of the network. The class defines the 8〇2 n wireless network communication standard, which enables each network member to contact, communicate, and quickly exchange data, data, files, and audio and video information. ^ Due to the maturity of radio technology, the wireless network is booming and the 'on the other hand'. With the development of various applications, users have more and more demand for bandwidth, so how effective is it. The improvement of bandwidth has become one of the focuses of modern information industry research and development. In addition to the advancement of physical layer transmission speed, the design of media access control also accounts for a large factor. Since each network member transmits data via a wireless medium, 4Γ376123, it is necessary for each network member to comply with the same agreement to coordinate the access of each network member to the media. Generally, the access mode used by the media access control (MAC) in the wireless network under the EEE802·11 standard is the "Distributed Coordination Function" (DCF) access mode. The station accesses the wireless medium according to a carrier sensing multiple access/collision avoidance (CSMA/CA) protocol.

According to the above agreement, when a station A sends data, it will first detect whether the current wireless medium is idle (idle, that is, to detect whether other stations are already transmitting in the wireless medium). If the wireless medium is in an idle state, the station A randomly selects a backoff time to start the countdown and continuously detects the wireless medium. When the back time of the backoff time ends, if the wireless medium is still interposed, it represents the station A. Have access to the media, so you can start to send out its frame. On the other hand, if station A has already occupied the wireless medium before the end of the countdown (think that other stations have started to send the frame), then station A will continue to count the number of days until the station a detects the wireless medium again. In the idle state, since A Jing will open 35 again to carry out the retreat of the time limit, in order to hold the right to access the wireless media. In other words, these stations are each independently selected by the independent machine - the time of retreat to coordinate the access rights of the wireless media to each other, so that the station with a shorter time will end the countdown and get the media first. Take the μ μ μ, the weight starts to send data; if there are more than two stations at the same time the countdown ends, then the user asks to send the data and causes a collision (c〇丨丨jsj〇n), which will start-retransmission mechanism. Obviously, according to the carrier detection multi-access/collision avoidance protocol 5 1376123 network performance, if the retreat 'but the wireless channel will have; otherwise 'if the retransmission time is re-transmission time, will also determine, the retreat time decision Will affect the entire wireless for a long time, then the probability of collision will be reduced too much idle time, resulting in a short decline in the use rate, the probability of collision will increase, increasing the effective use of the channel. In order to reduce the occurrence of repeated collisions, most of the current 8〇2 standard adopts the binary exponential concession mechanism (Binary Exp〇nentja|Ye (10), referred to as BEB), which will be used each time when the concession time is selected.

Select a value evenly, and its + cw is called the current contention window. When a station wants to start transmitting data, CW will set it to -min 纟CWmin, then start the retreat mechanism, and when the countdown ends, it will transmit If the data is successful, the next data will be prepared. If a collision occurs, the retransmission mechanism will be activated to multiply the value of cw and then start counting down. Since the selection range CW of the backoff time is doubled after each collision, the chance of selecting the same backoff time will be effectively reduced. If the collision continues, the value of CW continues to multiply until a maximum value CWmax is reached. After raising to the maximum value Cwmax, if the collision continues, the value of CW remains unchanged, and each time the collision is repeated, it is reversed again. If the number of retransmissions reaches the upper limit (R) or collision, the data will be discarded. Since the job mode that must be re-competitive after each piece of data is sent out is inefficient, therefore, in the new IEEE 802.11-2007 standard, a transmission opportunity (TranSmjssj〇n Opportunity, TX0P) is developed to improve network performance. The time value of Τχ〇ρ will be preset first. When a station grabs access to the wireless medium, it can access the wireless medium to the maximum TXOP for 6 1376123, thereby improving the performance of the network. With the advancement of network services, in addition to high-bandwidth, quality of service (QoS) guarantees are becoming more and more important. Therefore, an enhanced distributed channel access mode is defined in the IEEE 802.1 1-2007 standard. » EDCA), although it can provide four different types of services, but only the difference between different services can't provide a real guarantee. In addition, because each service also directly competes with wireless media, it will in some cases increase the chances of collisions and reduce channel utilization. SUMMARY OF THE INVENTION It can be seen from the above description that the media access control of the |EEE8〇211 wireless local area network accesses the media based on the carrier detection multiple access/collision avoidance protocol, generally adopts a binary exponential backoff mechanism (Binary Exponential BackQff). In order to reduce the probability of re-collision, but it will cause an increase in delay, it is difficult to achieve a good quality of service guarantee 'and no recommended parameter setting for the minimum competition window and transmission opportunities. In view of the above, the object of the present invention is to provide a wireless local area network with a service quality assurance fixed thief π 咳 乂 疋 疋 疋 疋 疋 疋 疋 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退 退The size of the stone machine is up to the optimal value, so that the wireless network to the best operating point to achieve the guarantee of service quality. The technical means used to achieve the above objectives, the fixed-view + L埤冯路视1^ withdrawal method with service quality assurance is performed by a communication device, which includes the following steps: Ding 1376123 Preparing to transmit data; The minimum competition window parameter parameter TXOP takes the -recession value in the [0, CWmin-1] interval; CVVmin and the transmission opportunity are randomly selected according to the uniform distribution and the idle value is reduced to the concession value when the idle time passes. Transfer data to zero unit, send TXOP time; one unit; can be continuous

Judging whether the data is successfully transmitted, ^ ^ w 7J is successful, ending the poor material transfer. The benefit storage process is performed by the above-mentioned preparation for transmitting data to transmit the next data, and vice versa;

Whether the number of J-failure failures exceeds - the number of retransmissions R is predetermined, if it is exceeded, then the next step is to make a step-by-step, and the same minimum contention window is used again. The number of steps CWmin is performed to perform the aforementioned random selection of the return value; and 7 the current data is discarded. And the above steps of preparing to transmit data are re-executed for the transmission of the next data. The aforementioned minimum contention window parameter CWmin and the transmission opportunity parameter TXO P are obtained by the execution-parameter method. The parameter setting method includes the following steps: - providing the probability parameter P of the collision, a given maximum delay time value D, - The given probability of loss parameter PL, the number of given communication devices n, the minimum transmission opportunity parameter TXOPmin with an initial value, and the variable maximum transmission opportunity parameter D X〇pmax of the initial value; according to the collision probability parameter p respectively calculate the probability τ and _ minimum competition window parameters of one station at random time transmission, · the minimum transmission opportunity parameter TXOPmin, the maximum transmission opportunity change 8 1376123 number TXOPmax, the number of retransmissions R, the random transmission rate 7 and the minimum competition window parameter Cwmin calculates a distribution of random variables d representing the time of two consecutive transmission opportunities; the probability that the random variable d is greater than the maximum delay D is less than the loss probability parameter PL, and the minimum transmission opportunity variable Tx〇Pmjn And the optimal transmission opportunity parameter TXOP when the collision probability parameter is p is found between the maximum transmission opportunity variable TXOPmax. Therefore, the present invention can obtain the most ideal minimum competition window parameter and transmission opportunity parameter by changing the collision probability parameter to p, and each station can cooperate with the fixed window retreat mechanism under these parameter settings, and the wireless network can be effective. The balance between service quality assurance and maximum channel usage. [Embodiment] Hereinafter, embodiments of the present invention will be described in conjunction with the drawings to make the contents of the present invention described above more particularly understandable. Before clarifying the fixed window retreat mechanism of the present invention, the conditions for satisfying the quality of service (Qos) are specifically defined: the wireless local area network includes a total of N communication devices, as shown in the first figure, The device comprises a wireless access point (Access P0int) (1) and a plurality of stations (7). The wireless access point (1) sets the same transmission opportunity time TXOP according to the number of communication devices n and the user service. Each station (7). Among them, the demand for service quality is: in the worst case. (ie, the wireless access point (1) and each station (2) continue to have data to be sent), a communication device transmits the τχ〇ρ hand twice in succession. The difference is a random variable..., the line network can ensure that the random variable d is greater than a predetermined 9 ^ 123. The probability of the maximum delay D is less than - the probability of this, p (d > D) S PL. Therefore, in order to meet the quality of service - demand will cause a delay increase. Controlling (4) the probability, avoiding frequent collisions, the fixed window retreat method of the present invention will apply a parameter setting::, the wireless access point (1) is calculated according to a parameter setting method - a good and fixed minimum competition View, defy the parameter CWmm and an optimal transmission opportunity parameter TXOP and pass 铉& as in & 乂 two parameters to each station (2), wireless access point (1) and each station ( 2) In this case, the smuggling < Λ ..., two > number S and each of them are operated according to the fixed window withdrawal method, which will enable the ... 0 regional network to effectively balance the service mouth guarantee And the best channel usage. The following is a detailed description of the fixed window retreat method of the present invention. Please refer to the first figure, including the following steps: Prepare to transmit data (1 01); Select the return value (102) to make the communication device according to - = small competition window The parameter one randomly selects a retreat access point (1) or a station (2) in a uniform distribution of gamma; the communication device may be a wireless reciprocal value (103), so that the communication device s 〇t ' when idle _ time slot That is, the retreat unit of the wireless network), the depreciation value is reduced by one unit; the unit = the transmission of the poor material (1〇4), so that the communication device reduces the value of the concession to zero. "(4) The data is transmitted at the beginning" It is possible to continuously transmit τχ〇ρ to judge whether the data is successfully transmitted (1〇5), and whether the data is detailed or not, and if the data is successful, the data transfer is completed. (1) Step (10) 1 is transmitted. The next-to-pen data, and vice versa, the next-step ^1376123, the number of failures is MM,. ^ 5 通讯 communication device] whether the number of failures exceeds a predetermined number of retransmissions R, - right-handed over, then - Otherwise use the same Small competition window parameters (4) to step (102); and ° - discard the data (1〇7), so that the communication device discards the current data when the number of failures exceeds the number of retransmissions R, and re-executes the step (1) For the transmission of the next data.

Base: In order to increase the channel utilization rate (IV), although the transmission opportunity time can increase the channel utilization rate, but it will cause the delay to rise and may not meet the service item f requirement, the transmission opportunity time must also be controlled to an optimal value. Find the maximum transmission opportunity time under the service quality requirement. Therefore, the optimization setting mentioned in the previous steps is to use the above parameter setting method to obtain the optimal minimum competition window parameter and transmission opportunity parameter. Therefore, please Referring to the third figure, the parameter setting method includes the following steps: Provide parameters such as P and, D, PL, N, TXOPmin and TXOPmax (2〇〇) 'provide a collision probability parameter p, a given maximum Delay time value D, a given loss probability parameter 卩匕, a given number of communication devices N, a variable minimum transmission opportunity parameter τΧ〇ΡίγΉ·η with an initial value, and a variable maximum transmission with an initial value The opportunity parameter TX〇Pmax; calculates the probability r of the random time transmission and the minimum contention window parameter CWmin(20l). In this embodiment, the collision probability parameter p is used. According to the public Ν-\Γη r αν. = one (2-Γχΐ-2ρ) (divination +_) - and Γ [〇-(2qing-P) + 冲_2冰丨-〆)] The probability of a station transmitted at random time and a minimum contention window parameter CWmin is obtained. The TXOP = (TXOPmin + TXOPmax)/2 (202) is taken, that is, the intermediate value of the minimum transmission opportunity parameter TXOPmin to the maximum transmission opportunity parameter TXOPmax is obtained. As the current transmission opportunity parameter TXOP; calculating the distribution of the random variable d (203), the distribution of the random variable d is calculated according to the number of retransmissions R, the random transmission probability τ, the minimum contention window parameter CWmin, and the current transmission opportunity parameter TXOP. The distribution of random variables d can be calculated by reference [1][2] ([1] : A. Banchs, P. Serrano and A' Azc.orra, End-to-end delay analysis and admission control in 802.1 1 DCF WLANs, ''

Compute Communication, vol. 29, issue 7, pp 842-584, April 2006. [2] : Hai L. Vu and Taka Sakurai, “Accurate Delay Distribution for IEEE 802.1 1 DCF, ” IEEE Communication Letters, vol. 10, no 4, 2006.); Calculate P(d>D)(204)' is to calculate the probability p(d>D) that the random variable d is greater than the maximum delay D; to determine whether P(d>D) is greater than PL(205)' If the probability p(d>D) is greater than the loss probability parameter PL, it indicates that the current TXOP is too large and needs to be small, so that the aforementioned maximum transmission opportunity parameter TXOPmax is equal to the value of the current transmission opportunity parameter TXOP and thus the maximum transmission opportunity parameter TX 〇pmax returns to step (202) to continue searching for the ideal TXOP, otherwise to proceed to the next step; to determine whether the difference between P(d>D) and PL is small enough (2〇6), that is, to judge P(d>D Whether the difference from PL is less than a specific value, if it is less than the specific value 12 1376123, then the step - step 'is the opposite. It means that the current # τ χ〇 ρ is too small to be large, so the minimum transmission opportunity is recorded TX 〇 Pmjn is equal to At present, the value of the opportunity parameter TXOP is transmitted and returns to the step '(2〇2) to continue to find the ideal TXOP; The optimal TXOP value (207) is obtained, and τ χ〇 ρ at this time is the optimal transmission opportunity parameter ΤχΟ ρ when the collision probability parameter is ρ.

The foregoing steps GO4) to GO7) are based on the premise that the probability that the random variable is greater than the maximum delay D is less than the loss probability parameter pL, and an optimal transmission is found between the minimum transmission opportunity parameter 及_ and the maximum transmission opportunity parameter. Opportunity parameter TXOP. When applied to an actual wireless local area network system, the setting of the collision probability parameter P value may be performed by the user after using the PL value of the missing probability parameter, and using the parameter setting method to find each p in a suitable range of P values. The optimal transmission opportunity parameter τχορ corresponding to the value is calculated according to the obtained transmitter b parameter TXQρ, and the optimal channel P value is obtained after comparing the channel usage rate '. For example, given the aforementioned maximum delay parameter D = 4 〇 ms, the loss probability parameter PL_〇·01, the number of retransmissions R = 6, the collision probability parameter p is 0.05 0.45, and the operation is performed in the above step, and the fourth figure is obtained. The relationship between the collision probability and the channel usage rate shows that the transmission probability parameter TXOP is the best when the collision probability parameter P is equal to 〇·17, regardless of the number of communication devices N. value. In practical application, the optimal p value corresponding to each of the missing probability parameter values can be calculated in advance, and then the matching combination of the parameters is stored in the wireless access point (1) to enable wireless The access point (1) is determined by the PL·reset 13 1376123, i.e., the matching P group according to the stored parameters corresponds to obtain the best P value. As described above, the operation procedure of the parameter setting of the wireless access point (1) in the continuous system can be referred to the fifth figure: when the administrator sets the required loss rate parameter PL of the wireless access point (1) _ ), the wireless access point (1) will get the best P value according to the matching combination of the stored parameters _) 'Wireless access point (1) according to the best tp value can be used to obtain the best using the above parameter setting method The minimum contention window parameter (4) - and the transmission opportunity parameter TXOP, and the obtained parameters are transmitted to each # (7) setting, after which all communication devices (ie, wireless access point and station) perform data transmission according to the aforementioned fixed window method; When a new station is added or the original station leaves (302), due to the change in the number of communication devices, the wireless access point (1) will recalculate the optimal τχ〇ρ盥CW coffee (303) according to the aforementioned parameter setting method; Wireless access point (1) (4) When the new station joins, causing the new τχορ to be too small to transmit the maximum packet (304), the Xiao wireless access point (1) will reject the joining of the new station (305), otherwise the wireless access point (1) will be the club weight The best τχορ and CWmin are set to each station (306) 〇° The above channel usage rate is calculated by a formula: where Te is the time of the wireless network basic unit s|〇t; The time required for the success of transmitting τχορ; Tc is the waste of the collision; Spt "the probability of transmitting at least one station in a certain slot is outside = 1__ (u.

In the case of a certain slot towel, at least one station transmits ‘PS ps_ ητ(1-τ)Ν~χ好/, 有—the probability of station transmission, ie i-U-f. 1376123 The fixed window retreat of the present invention is different from the commonly used binary exponential method. After determining the optimal size of the small competition window CWmin, the fixed value is no longer changed, and there is a Q〇S. Under the request, the channel utilization rate of the present invention and the binary index retreat mechanism can be compared with the fifth figure of the moxibustion book & eve, it can be clearly seen that the effect of the fixed window retreat mechanism of the present invention is greater than Standard binary index concession mechanism. If the experiment is further carried out with different communication, the difference between the two will be more obvious.

In summary, the present invention proposes a fixed window retreat mechanism for the media access control layer of the |EEE8Q211 wireless network, and there is no parameter setting for the minimum contention window and transmission opportunity in the prior art field. The method, therefore, the present invention also provides a parameter setting method, regardless of the number of communication devices, as long as the given probability of loss & pL, the best collision probability p can be found, and the best minimum competition window CWmin is obtained accordingly. With the transmission opportunity parameter TXOP, the communication device of the wireless local area network operates in conjunction with the fixed window retreat mechanism under these ideal parameter settings, which can effectively balance the quality of service guarantee (Q〇S) and the maximum channel utilization rate (Channel Utilization). [Simple description of the diagram] The first picture: a schematic diagram of the member structure of a communication device of a wireless network. The first figure: the flow chart of the fixed window retreat method of the present invention, the second figure · The parameter setting process of the fixed window retreat method of the present invention. The fourth picture: the relationship between the collision probability and the channel utilization rate of the present invention under the number of different communication devices . 15 1376123 Fig. 5 is a flow chart showing the operation of parameter setting using the wireless access point (1) of the present invention. Figure 6: Comparison of the present invention and the binary exponential depreciation method in terms of collision probability and channel usage. [Main component symbol description] (1) Wireless access point (2) Platform

16

Claims (1)

1376123 VII. Patent application scope: Concession 1. A fixed wireless method for wireless local area network with service quality assurance is performed by a communication device, which includes the following steps to prepare to transmit data; according to an optimal minimum competition window parameter CWmin And the transmission opportunity parameter takes a TXOP in the [〇, cWmin-l] interval of the towel step, and J1< 随机 randomly selects the retreat value according to the uniform distribution; one unit; can be continuously continuous for one time slot time Reduce the return value After the return value is reduced to zero unit, the data transmission is started, and the TXOP time is transmitted; whether the data is successfully transmitted is judged, if successful, the data transmission is ended and the above-mentioned preparation data transmission step is executed again to transmit the next data, otherwise, the next step is performed. Determining whether the number of failures exceeds a predetermined number of retransmissions R, and if so, proceeding to the next step 'again, repeating the step of randomly selecting the retreat value using the same minimum competition window parameter CWmm; and discarding the current data and re-executing the foregoing Prepare the transfer data step for the next data transfer. 2. The fixed-area retreat method for the wireless area network with service quality assurance according to the scope of claim 1 of the patent application, the minimum contention window parameter (4) 2 and the transmission opportunity record TXQP are obtained by the parameter setting method, " The parametric parameter attack method includes the following steps: 1 2 supply-collision probability parameter P, a given maximum delay time value D, σ-determined loss probability parameter PL, a given number of communication devices N, an initial value The minimum transmission opportunity is less than TXOPmin and the initial value of 17 variable large transmission opportunity parameter TX0Pmax; according to the collision probability parameter P respectively calculated - the probability of station transmission at random time r and t ^ Small competition window parameter CWmin; take the most) transmission opportunity parameter TXOPmin to the maximum conveyor. The parameter TXOPrnax's mid-p bow gentleman 仏i B value as the current transmission opportunity parameter TXOP; Hi transmission number R, random transmission probability r, minimum The transmission window parameter TXOP of the competition window parameter and the target m calculates a random variable k distribution representing the time of two consecutive transmission opportunities; with the variable d being greater than the maximum delay D Reference rate is less than the probability of loss for the lead to raise the minimum and maximum transmission opportunity variables TX〇Pmiri variable transmission opportunities TXQPmax find the material - the best money transfer (iv) will τχ〇ρ. The patent scope of the towel is 帛2帛, the line area network has service items: the guaranteed fixed window retreat method, the above-mentioned "the probability that the random variable d is greater than the most = D is less than the loss probability" and the minimum transmission The second step of the variable = T_min and the maximum transmission opportunity variable is to find the best transmission opportunity parameter. The steps include the following steps: P(d>D), which calculates the random variation probability P(d>D). ; machine-number d is greater than the maximum delay D is due to legacy!::(d>D) is greater than the loss probability parameter pl, if P((4)) is equal to the target H parameter PL' material maximum material opportunity parameter TXOPmax The transmission opportunity parameter D0X transmission opportunity parameter TX〇Pmin to the second=to the above “take the minimum intermediate value as the current transmission opportunity parameter Ding Qingyan to proceed to the next step; τχ〇ρ” step, and vice versa 18 1376123 judgment Whether the difference between P(d>D) and PL is small enough, that is, whether the difference between p(d>D) and the missing probability parameter PL is less than a specific value, and if it is smaller than the specific value, the next step is performed, and vice versa. Then the minimum transmission opportunity parameter TXOPmin is equal to the current transmission opportunity Number τχ〇ρ value and return to the previous "fetch the minimum transmission parameters TXQPmin opportunity to TX〇PmaX maximum transfer parameter opportunity median value as the current transmission opportunity TXOP parameters" step The best TXOP value is obtained. When the collision probability parameter is p, the transmission opportunity parameter TXOP at this time is the optimal transmission opportunity parameter TXOP. Eight, schema. (such as the next page) 19