CN1917708A - Reception contro method based on measurement and QoS in broadband radio access system - Google Patents

Abstract

The present invention belongs to the field of admission control methods in broadband wireless access systems, and is characterized in that it allows access to best-effort services when judging service types; non-application grant services with the first priority and non-real-time services with the third priority The polling service is judged according to whether the sum of the actual bandwidth of the service and the bandwidth used by the system is less than the given bandwidth of the system; for the real-time polling service of the second priority, it is first screened according to the actual bandwidth of the service as described above, and secondly, whether it is GPC mode judgment, if it is GPC mode, judge according to service delay. The present invention simultaneously adopts an admission control strategy combining actual bandwidth and delay, improves real-time business service quality, and optimizes system performance.

Description

Admission Control Method Based on Measurement and QoS in Broadband Wireless Access System

technical field

The invention relates to an admission control algorithm in a broadband wireless access (BWA) system based on IEEE 802.16, belonging to the technical field of wireless communication.

Background technique

The IEEE 802.16 standard is the air interface specification of the wireless metropolitan area network, which has very important significance and broad application prospects in the future broadband wireless access. The 802.16 standard not only allows non-line-of-sight connections and supports large-capacity users, but also provides carrier-level QoS guarantees and fully supports the application of rich multimedia services such as voice, video, and data.

An effective QoS guarantee system must allocate wireless resources reasonably, especially for the complex environment of wireless communication environment, the QoS system of the wireless access system faces many challenges. The 802.16 standard implements a series of mechanisms at the MAC layer. Admission control is an important strategy in the QoS guarantee system of broadband wireless access system. Through admission control, the system can estimate system resources and service characteristics before formal communication, allow service access under the premise that it can meet the service QoS of the new access and not affect the QoS of the service already accessed, and refuse admission if it does not meet the requirements. conditions of business.

The IEEE 802.16 system has four different service levels of services, namely unsolicited grant service (UGS), real-time polling service (real-time polling service, rtPS), non-real-time polling service (non-real-time polling service) , nrtPS), best effort service (best effort service, BE).

The existing admission strategy is a bandwidth-based admission method, and its specific implementation process is as follows:

Given the available bandwidth B of the system, assume that when a new service arrives, the bandwidth already used by the system is _Bused , and the bandwidth required by the UGS service is b _UGS , the minimum bandwidth required by the rtPS and nrtPS services is b _min , and the maximum bandwidth is b _max .

If the new service is a UGS service, then if B _used +b _UGS < B, allow the UGS service to access the system and allocate the reserved bandwidth of b _UGS , otherwise reject it;

If the new service is an rtPS service, if B _used +b _max < B, allow service access and allocate the reserved bandwidth of b _max , otherwise reject;

If the new service is an nrtPS service, if B _used +b _min < B, allow service access and allocate b _min reserved bandwidth, otherwise reject;

All BE services are allowed to be accessed, but reserved bandwidth is not allocated.

Since the required bandwidth of the rtPS and nrtPS services is distributed between (b _min , b _max ), it is not very reasonable to use the maximum or minimum bandwidth to judge whether to accept each time. For example, for the rtPS service, if the incoming rtPS service does not require the maximum bandwidth, but the access is judged according to the maximum bandwidth, the blocking rate of the rtPS service will be relatively high, which will lose fairness compared with other services.

At the same time, for real-time rtPS services, the existing admission control method only adopts the method of bandwidth control, and does not consider parameters such as delay and packet loss rate of real-time services. Therefore, the service quality of real-time services cannot be guaranteed. A good admission algorithm needs to comprehensively consider business service quality parameters such as bandwidth, delay, and packet loss rate, so as to provide QoS guarantee for the communication system.

Contents of the invention

The purpose of the present invention is to propose an admission control algorithm based on measurement and QoS guarantee applied in the IEEE 802.16 broadband wireless access system, which can effectively reduce the packet loss rate of real-time services without affecting the utilization rate of system bandwidth resources. Improve service quality and get better system performance.

The present invention is characterized in that it contains the following steps in sequence:

Step (1), when the new business arrives, the base station of the broadband wireless access system based on IEEE 802.16 judges the type of the new business, and ranks the non-application granting business therein as the first priority, and polls the business column in real time It is the 2nd priority, and the non-real-time polling service is listed as the 3rd priority, according to the following steps for different business types;

Step (2), for the best-effort service, the base station allows access, then go to step (6);

Step (3), for non-applicable and granted services, the base station judges whether the given system available bandwidth B is greater than the sum of the system's already used bandwidth B _used and the bandwidth B _UGS required by the new service, if it is larger than the system's given available bandwidth , the access is allowed, go to step (6), otherwise, the access is denied;

Step (4), for the non-real-time polling service, the base station performs the following steps:

Step (4.1), measuring the actual bandwidth b _m of the service;

Using the time window measurement method, the steps are as follows:

作为估计带宽；②当接入新流时，更新估计带宽，并重启时间窗；③当

超过估计带宽时，以 作为估计带宽；④当有一流离开时，更新估计带宽，并重启时间窗；估计带宽的更新公式如下：Divide a time window into 10 sampling periods, and calculate the average bandwidth usage of each type of service in each period (sampling time is S), update the value of the estimated bandwidth in the following four cases: ① When the time window ends, take the largest

As the estimated bandwidth; ② When a new stream is connected, update the estimated bandwidth and restart the time window; ③ When

When the estimated bandwidth is exceeded, the As the estimated bandwidth; ④ When a class leaves, update the estimated bandwidth and restart the time window; the update formula of the estimated bandwidth is as follows:

为S中的平均使用带宽；Wherein, b _m is the updated estimated bandwidth, that is, the actual measured bandwidth of the service;

is the average used bandwidth in S;

r ^p is the generation rate of business information;

β is the bandwidth reduction factor. When a flow ends and leaves the network, the estimated value of the used bandwidth needs to be reduced. The reduced bandwidth is roughly equal to the equivalent bandwidth of the flow. Different values can be determined for β according to different flows. In order to simplify the algorithm, Can take β=1;

Step (4.2), judging whether the available bandwidth B of the given system is greater than the sum of the bandwidth used by the system and the actual bandwidth b _m of the service, if the available bandwidth B of the system is large, allow access, and turn to step (6) , otherwise deny access;

Step (5), for the real-time polling service, the base station performs the following steps:

Step (5.1), measuring the actual bandwidth b _m of the service, the measurement method is the same as step (4.1);

Step (5.2), judging whether the available bandwidth B of the system is greater than the sum of the bandwidth B _used and the actual bandwidth used by the system;

If the system available bandwidth B is large, then enter step (5.3);

Otherwise, access is denied;

Step (5.3), judge whether described business is GPC mode, and this mode is meant that base station allocates setting bandwidth for each connection;

If it is GPC mode, then judge:

p _i <P _i No, when p _i <P _i allows access, otherwise rejects access,

Wherein, p _i is the packet loss rate of the service i calculated according to the following formula:

${\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Ts</span>}<span\; class="notranslate">\; )</span>}$

λ _i is the Poisson parameter of the arrival of the business packet, a known value obtained by measurement;

μ _i is the average service time of the business package, a known value obtained by measurement;

D _i is the maximum delay allowed by service i, the set value;

Ts is frame time compensation, set value;

p _i is the maximum packet loss rate set according to the service quality index;

If it is not GPC mode, follow the next step;

Step (5.4), is carried out by GPSS mode, and this mode refers to that the base station allocates the set bandwidth for each terminal, and the terminal redistributes the bandwidth among its several connections;

The base station determines whether W ₂ <D or not,

${\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="W"\; filenames="A20061011277500132.png"\; state="\{\{state\}\}">W</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>}$

T ₁ is the time to wait for the service desk to be vacated , λ _i is the measured Poisson parameter of the arrival of the i-th priority packet, h _i is the average service time of the i-th priority packet, and σ _i ² is the variance of the service time;

ρ ₁ = λ ₁ h ₁ , ρ ₂ = λ ₂ h ₂ , λ, h are as mentioned above;

D is the maximum delay allowed by the real-time polling service, the set value;

W ₂ is the delay of the real-time polling service;

Judgment result, if W ₂ <D, access is allowed, otherwise, access is denied;

In step (6), the base station allocates reserved bandwidth.

The present invention introduces a measurement mechanism to obtain relevant system parameters in real time, and adopts an admission control strategy combining bandwidth and delay to improve real-time business service quality and optimize broadband wireless access system performance.

Description of drawings

Figure 1. QoS guarantee system of IEEE 802.16 system;

Figure 2. The flow chart of the existing admission control algorithm;

Fig. 3. the admission control algorithm flow chart that the present invention proposes;

Figure 4. Comparison of the blocking rate before and after the introduction of the measurement mechanism: Figure 4a shows the blocking rate of each service in the existing algorithm, where the straight line is the blocking rate curve of the non-application grant service, the asterisk line is the real-time service blocking rate curve, and the dotted line is Non-real-time service blocking rate curve; Figure 4b shows the service blocking rate of the algorithm after introducing the measurement mechanism, where the straight line is the non-application grant service blocking rate curve, the asterisk line is the real-time service blocking rate curve, and the dotted line is the non-real-time service blocking rate curve rate curve;

Figure 5. Comparison of rtPS service packet loss rate before and after introducing delay admission control;

Fig. 6. Comparison of system resource utilization before and after adopting the present invention.

Detailed ways

In order to realize the technical problem to be solved in the present invention, the present invention adopts a two-layer admission control system based on bandwidth and delay that introduces a measurement mechanism, including the following steps:

For all services, the admission judgment is first made based on the bandwidth. If the sum of the bandwidth used by the system and the bandwidth of the new service is less than the total bandwidth that the system can provide, the service access is allowed, otherwise it is rejected. In order to ensure real-time performance, the bandwidth of the rtPS and nrtPS services is obtained by using a measurement method, and the bandwidth admission judgment is performed by measuring the bandwidth.

For the real-time rtPS service, after the bandwidth judgment in (1) is performed, the delay judgment is performed. According to the parameters of the measured business flow, the relevant methods of queuing theory are used to model, and the delay or packet loss rate of the rtPS business is measured, and compared with the corresponding service quality parameters, it is judged whether to allow admission. That is, if the delay or packet loss rate indicators of real-time services meet the quality of service requirements, access is allowed, otherwise access is denied.

Specific embodiments of the present invention will be further described below.

Fig. 3 is the flow chart of admission control algorithm of the present invention, specifically described as follows:

After the arrival of a new service, the admission control judgment based on the bandwidth is firstly performed. Given the bandwidth B available to the system, assume that when a new service arrives, the bandwidth already used by the system is _Bused , and the bandwidth required by the UGS service is b _UGS , and the measured actual bandwidth of the rtPS and nrtPS services is b _m . The admission control module first judges the service type.

If the new service is a UGS service, then if B _used +b _UGS < B, allow the UGS service to access the system and allocate the reserved bandwidth of b _UGS , otherwise reject it;

If the new service is an rtPS service, if B _used + b _m < B, start to judge the delayed acceptance, otherwise reject;

If the new service is an nrtPS service, if B _used + b _m < B, allow access and allocate the reserved bandwidth of b _m , otherwise reject;

All BE services are allowed to be accessed, but reserved bandwidth is not allocated.

For the rtPS service entering the delayed admission judgment in (2), the admission judgment is performed according to the delayed admission control algorithm.

802.16 MAC has two modes, one is GPC (Grant Per Connection) mode, that is, the base station allocates a certain bandwidth for each connection, and the other is GPSS (Grant Per Subscriber Station) mode, that is, the base station allocates a certain amount of bandwidth for each terminal. The terminal redistributes the bandwidth among its several connections.

For the rtPS service of the entry delay admission judgment in (2), it is first judged which MAC mode it belongs to.

If it is GPC mode, adopt the delay control strategy based on GPC mode.

The admission judgment condition is: if p _i <P _i , then allow rtPS service access, otherwise reject it.

Wherein, _pi is the packet loss rate of service i, and P _i is the maximum packet loss rate as a service quality index, which is a fixed parameter.

The calculation process of the packet loss rate p _i of business i is as follows:

Assuming that the arrival of business packets obeys the Poisson distribution, and the service time of the packets obeys the negative exponential distribution, then the cache at the SS end is modeled approximately as an M/M/1 queue.

The measured Poisson parameter of the arrival of the service packet is λ _i , and the negative exponential distribution parameter is μ _i .

According to the conclusion of queuing theory, the distribution of waiting time W of packets is:

$\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; w</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\frac{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>\mathrm{<span\; class="notranslate">\; w</span>}}$

Then the packet loss rate of service i can be calculated as:

${\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; ></span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Di</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Ts</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; G</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Di</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Ts</span>}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; Di</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Ts</span>}<span\; class="notranslate">\; )</span>}$

Among them, D _i is the maximum delay that service i can tolerate, and Ts is the framing time compensation.

Then the packet loss rate p _i of service i can be obtained, and then judge whether the service is accessed or not according to the judgment expression p _i <P _i of delay acceptance.

If it is in GPSS mode, the delay control strategy based on GPC mode is adopted.

The condition of admission judgment is: if W ₂ <D, access is allowed, otherwise, it is rejected.

Among them, W ₂ is the delay of the rtPS service, and D is the maximum delay allowed by the service.

The calculation process of _W2 is as follows:

In GPSS mode, the scheduling of the SS side meets the priority criterion, that is, among the four levels of service queues on the SS side, packets with high priority are sent to the wireless channel first, and packets with low priority are sent only when the high-priority queue is empty. You can send.

Modeled as a priority queue. Among them, UGS is the first priority, rtPS is the second priority, and nrtPS is the third priority.

Suppose the measured Poisson parameter of the arrival of the i-th level packet is λ _i , the average service time is h _i , the variance is σ _i ² , the total traffic volume of the system is ρ=Σλ _i h _i , ρ<1, and W _i is The waiting time of the i-th level packet.

According to the priority queuing theory, assuming that the waiting time for the service desk to be vacated is T ₁ , then there is W ₁ =T ₁ /(1-ρ ₁ ), and the average waiting time of the k-th class packet is obtained by recursion:

${\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; /</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

$<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>\underset{\mathrm{<span\; class="notranslate">\; i</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>}$

Therefore, the average waiting time for rtPS packets is:

${\mathrm{<span\; class="notranslate">\; <figure-callout\; id="2"\; label="W"\; filenames="A20061011277500132.png"\; state="\{\{state\}\}">W</figure-callout></span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>}$

in, ${\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="A20061011277500122.png,A20061011277500132.png"\; state="\{\{state\}\}">T</figure-callout>}}_1=\mathrm{\&Sigma;}\frac{{\mathrm{\&lambda;}}_i}{2}({\mathrm{<figure-callout\; id="2"\; label="h\; i"\; filenames="A20061011277500132.png"\; state="\{\{state\}\}">h</figure-callout>}}_i^{}+{\mathrm{\&sigma;}}_i^2),$ ρ _i =λ _i h _i .

Then the delayed admission judgment expression in GPSS mode is

W ₂ <D

Then the delay of the rtPS service in the GPSS mode can be obtained, and then judged according to the delay acceptance expression W ₂ <D, if the condition is met, the access will be accepted, otherwise, it will be rejected.

In order to compare the effects before and after the present invention is implemented, the algorithm of the present invention is simulated with Matlab, and the simulation parameters adopt B=2000kbps, b _UGS =32kbps, _bmin =32kbps, _bmax =32kbps, and the Poisson parameters of three kinds of business flows are identical , and the values of λ range from 1 to 10, and the service time distribution parameters of the three services are all 3.05. Assume that the actual measured rtPS and nrtPS service bandwidths are evenly distributed on (32, 128) kbps, and the actual measured service delay is positively correlated with the size of λ, and obeys the exponential distribution.

Figure 4 shows the comparison of the blocking rate of each service before and after the introduction of the measurement mechanism. It can be seen that the original admission policy uses the maximum bandwidth to judge whether to accept the rtPS service, so that the blocking rate of the rtPS service is greater than that of other services, and the measurement mechanism is introduced. Finally, the blocking rate of the rtPS service is balanced with other services. The audio and video service corresponding to rtPS occupies an important position in the future wireless access network, so the introduction of the measurement mechanism can effectively improve the fairness of the service.

Figure 5 shows the comparison of the packet loss rate of rtPS services before and after delay admission is adopted. After introducing the delay-based admission control strategy for rtPS services, it can be seen that the packet loss rate of real-time services decreases significantly, and with the increase of system load , the packet loss rate remains stable. At the same time, due to the limitation of delay control, the blocking rate of the rtPS service will increase.

Fig. 6 is a comparison of system bandwidth resource utilization before and after adopting the present invention. Since the admission policy that integrates bandwidth and delay control can effectively reduce the packet loss rate of real-time services and ensure the QoS of services, it also increases the blocking rate of real-time services to a certain extent, and the introduced measurement mechanism can reduce the real-time service rate. Therefore, we compare the system resource utilization rate of the original admission algorithm in the 802.16 system and the new algorithm proposed by the present invention. In Figure 6 it can be seen that using

After the algorithm of the present invention, the resource utilization rate of the system does not change much, and is basically equal to the original system resource utilization rate, but at the same time, the service quality of the real-time rtPS service is obviously improved, so the performance of the system is improved.

Claims (1)

1. An admission control method based on measurement and QoS in a broadband wireless access system, characterized in that it contains the following steps in sequence: Step (1), when the new business arrives, the base station of the broadband wireless access system based on IEEE 802.16 judges the type of the new business, and ranks the non-application granting business therein as the first priority, and polls the business column in real time It is the 2nd priority, and the non-real-time polling service is listed as the 3rd priority, according to the following steps for different business types; Step (2), for the best-effort service, the base station allows access, then go to step (6); Step (3), for non-applicable and granted services, the base station judges whether the given system available bandwidth B is greater than the sum of the system's already used bandwidth B _used and the bandwidth B _UGS required by the new service, if it is larger than the system's given available bandwidth , the access is allowed, go to step (6), otherwise, the access is denied; Step (4), for the non-real-time polling service, the base station performs the following steps: Step (4.1), measuring the actual bandwidth b _m of the service; Using the time window measurement method, the steps are as follows: Divide a time window into 10 sampling periods, and calculate the average bandwidth usage of each type of service in each period

The sampling time is S, and the value of the estimated bandwidth is updated in the following four cases: ① When the time window ends, take the largest

As the estimated bandwidth; ② When a new stream is connected, update the estimated bandwidth and restart the time window; ③ When

When the estimated bandwidth is exceeded, the

As the estimated bandwidth; ④ When a class leaves, update the estimated bandwidth and restart the time window; the update formula of the estimated bandwidth is as follows: Wherein, b _m is the updated estimated bandwidth, that is, the actual measured bandwidth of the service; is the average used bandwidth in S; r ^p is the generation rate of business information; β is the bandwidth reduction factor. When a flow ends and leaves the network, the estimated value of the used bandwidth needs to be reduced. The reduced bandwidth is roughly equal to the equivalent bandwidth of the flow. Different values can be determined for β according to different flows. In order to simplify the algorithm, Can take β=1; Step (4.2), judging whether the available bandwidth B of the given system is greater than the sum of the bandwidth used by the system and the actual bandwidth b _m of the service, if the available bandwidth B of the system is large, then allow access, and turn to step (6 ), otherwise deny access; Step (5), for the real-time polling service, the base station performs the following steps: Step (5.1), measuring the actual bandwidth b _m of the service, the measurement method is the same as step (4.1); Step (5.2), judging whether the available bandwidth B of the system is greater than the sum of the bandwidth B _used and the actual bandwidth used by the system; If the system available bandwidth B is large, then enter step (5.3); Otherwise, access is denied; Step (5.3), judge whether described business is GPC mode, and this mode is meant that base station allocates setting bandwidth for each connection; If it is GPC mode, then judge: p _i <P _i No, when p _i <P _i allows access, otherwise rejects access, Wherein, _pi is the packet loss rate of the service i, which is calculated as follows: ${\mathrm{<span\; class="notranslate">\; p</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}}{\mathrm{<span\; class="notranslate">\; e</span>}}^{<span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; \&mu;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&lambda;</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; D.</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; Ts</span>}<span\; class="notranslate">\; )</span>}$ λ _i is the Poisson parameter of the arrival of the packet, which is a known value obtained by measurement; μ _i is the complex exponential distribution parameter of the service time of the packet, a known value obtained by measurement; D _i is the maximum delay allowed by service i, the set value; Ts is frame time compensation, set value; p _i is the maximum packet loss rate set according to the service quality index; If it is not GPC mode, follow the next step; Step (5.4), is carried out by GPSS mode, and this mode refers to that the base station allocates the set bandwidth for each terminal, and the terminal redistributes the bandwidth among its several connections; The base station determines whether W ₂ <D or not, ${\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; T</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}}{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&times;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; \&rho;</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; )</span>}$ T ₁ is the time to wait for the service desk to be vacant, $T_1=\mathrm{\&Sigma;}\frac{{\mathrm{\&lambda;}}_i}{2}(h_i^2+{\mathrm{\&sigma;}}_i^2),$ λ _i is the measured Poisson parameter of the arrival of the i-th priority packet, h _i is the average service time of the i-th priority packet, and σ _i ² is the variance of the service time; ρ ₁ = λ ₁ h ₁ , ρ ₂ = λ ₂ h ₂ , λ, h are as mentioned above; D is the maximum delay allowed by the real-time polling service, the set value; W ₂ is the delay of the real-time polling service; Judgment result, if W ₂ <D, access is allowed, otherwise, access is denied; In step (6), the base station allocates reserved bandwidth.

Images