CN103618674A - A united packet scheduling and channel allocation routing method based on an adaptive service model - Google Patents

Abstract

The present invention provides a joint packet scheduling and channel allocation routing method based on an adaptive service model, which includes a sequentially connected cognitive information base, reasoning learning base, and adaptive routing scheduler, wherein: the cognitive information base is used for cognitive information Acquisition and modeling; reasoning learning library is used for prediction and reasoning of resource allocation sub-decisions; adaptive routing scheduler is used for joint packet scheduling and channel allocation to make global optimal decisions and control the allocation of entire network resources. The present invention makes a globally optimal routing decision through a joint packet scheduling and channel allocation algorithm, which can significantly improve the service quality of different multimedia services in the mobile cognitive wireless network, and highly dynamically and adaptively adjust the mobile cognitive wireless network Decision models in the protocol. The separation of cognitive information, decision-making model and routing makes the algorithm easier to expand and maintain, thus providing impetus and support for a unified standard mobile cognitive wireless network QoS routing method.

Description

Joint Packet Scheduling and Channel Assignment Routing Method Based on Adaptive Service Model

technical field

The invention belongs to the field of computer technology, in particular to a routing method for joint packet scheduling and channel allocation based on an adaptive service model in a mobile cognitive wireless network environment, and specifically relates to a joint packet scheduling and channel allocation based on an adaptive service model routing method.

Background technique

Cognitive radio is a new technology extended on the basis of software radio, which was first proposed by Dr. Joseph Mitola of the Swedish Royal Institute of Technology in 1999. The emergence of cognitive radio technology has broken the traditional fixed spectrum allocation system. It can sense the current wireless environment in real time, find out the unused licensed spectrum, and dynamically use these "spectrum holes" through opportunistic spectrum access. Realize the secondary utilization of the licensed spectrum, improve the utilization rate of the licensed spectrum, and effectively alleviate the increasingly tense problem of wireless spectrum resources. In cognitive wireless networks, there are two types of users, primary users (Primary User, PU) and secondary users (Secondary User, SU). The primary user is an authorized user assigned a fixed frequency band, and has priority to use the spectrum. Secondary users are also called cognitive users, which are unlicensed users who dynamically access the licensed spectrum and use "spectrum holes" to communicate without interfering with the primary user's communication.

Traditional cognitive wireless networks provide a single best-effort service for data transmission services. However, with the continuous emergence of some multimedia services such as video-on-demand and IP telephony, different levels of requirements are put forward for network quality of service (QoS). For example, real-time voice services require small transmission delay and jitter, and certain errors are allowed; while data transmission services require high reliability and general delay requirements. In addition, in mobile cognitive wireless networks, the mobile speed, mobile mode, and access spectrum of each node are random, which may cause changes in network topology and channel bandwidth at any time. The inherent dynamics of the mobile wireless cognitive network make it difficult to meet the service quality requirements of different services. How to provide different services according to different business QoS requirements has always been the difficulty that mobile cognitive wireless networks are committed to solving, and routing selection based on the cognitive information of each layer is the key technology to solve QoS.

After searching the existing technical documents, it is found that the current research on the QoS routing method for the mobile cognitive wireless network is still in its infancy, and there is no recognized model. He Qing and Zhou Huaibei (He Qing,ZhouHuaibei.Research on the routing algorithm based on Qos requirement forcognitive radionetworks[C].2008International Conference on Computer Science and SoftwareEngineering.2008) according to the channel capacity and delay, the joint network layer and MAC The layer adopts the idea of minimum spanning tree to find the best routing path, which is used to reduce the delay in the whole routing data forwarding process. This QoS routing algorithm tries to find a balance point between channel capacity and delay, and finally obtains a routing selection with the best QoS performance. However, the QoS performance of the network does not only refer to a single performance such as delay, and different applications have different performance requirements in different occasions. In addition, the cognitive information and control information are fused together in the QoS routing mechanism, which makes the protocol difficult to expand and maintain. Cognitive wireless networks must be able to adaptively allocate required network resources according to different business QoS requirements and different network environments, so as to achieve optimal allocation of network resources and improvement of network service quality. Therefore, mobile cognitive wireless networks need Routing methods for new scalable resources.

Contents of the invention

Mobile cognitive wireless networks must adaptively select routes according to dynamic network conditions, available spectrum, and different service QoS requirements, so as to improve user satisfaction and network resource utilization. Therefore, in order to complete the service transmission requirements in the shortest time, it is necessary to jointly consider packet scheduling and spectrum allocation based on service QoS for routing selection.

Based on the above understanding and the deficiencies in the prior art, the purpose of the present invention is to overcome the deficiencies of the existing mobile cognitive wireless network resource allocation mechanism, such as poor scalability and weak adaptive Due to the dynamic and changeable characteristics of wireless networks, a routing method based on joint packet scheduling and channel allocation based on adaptive service model in mobile cognitive wireless networks is proposed. The present invention uniformly models the requirements of different levels of services, network conditions, and available spectrum, and comprehensively considers scheduling and channel strategies when selecting routes to make global optimal routing decisions, thereby achieving self-adaptive satisfaction of different service QoS requirements, improving The purpose of the overall quality of service of the network.

According to the joint packet scheduling and channel allocation routing method based on the adaptive service model provided by the present invention, it includes sequentially connected cognitive information base, reasoning learning base and adaptive routing scheduler, wherein:

Cognitive information base is used to acquire and model cognitive information;

The reasoning learning library is used for prediction and reasoning of resource allocation sub-decisions;

The adaptive routing scheduler is used for joint packet scheduling and channel allocation to make global optimal decisions and control the allocation of the entire network resources.

Preferably, the cognitive information base includes service QoS requirement information, network node information, and available spectrum information, wherein:

Service QoS requirement information includes transmission delay, delay jitter, throughput and packet loss rate;

The network node information includes the neighbor information of the node;

Available spectrum information includes spectrum allocation, interference constraints, and benefits.

Preferably, the inference learning library includes a packet scheduling strategy module and a spectrum allocation strategy module, wherein:

The packet scheduling strategy module is used to adjust the weight of each service and each index in real time according to the current network changes, and dynamically adjust the position of each data packet in the waiting queue;

The spectrum allocation strategy module is used to predict the next round of spectrum allocation according to the current spectrum allocation situation.

Preferably, the adaptive routing scheduler is the controller of network resource allocation, controlling the scheduling, routing, and transmission channel of each data packet at the network node, and the adaptive routing scheduler is based on the information of the current network node and neighbor nodes Adaptively adjust the packet scheduling strategy and spectrum allocation strategy, find the most suitable data packet for scheduling from the candidate data packet set, and find the available channel from the predicted allocable spectrum, so that the joint packet scheduling and channel allocation will combine the current node's Data packets are transmitted from the optimal route.

Compared with the existing technology, the present invention has the following beneficial effects: the routing policy generated by joint packet scheduling and channel allocation in the shortest time is the global comprehensive optimal solution, which meets the QoS requirements of different services and reduces the cost of re-routing at the same time. Possibility, so it can greatly improve the service quality of the network. In addition, the self-adaptive service model is easier to expand and maintain through the separation of cognitive information, reasoning self-decision and joint scheduling.

Description of drawings

Other characteristics, objects and advantages of the present invention will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

FIG. 1 is a schematic diagram of an interaction principle between an adaptive service model and a traditional wireless network protocol.

Detailed ways

The present invention will be described in detail below in conjunction with specific embodiments. The following examples will help those skilled in the art to further understand the present invention, but do not limit the present invention in any form. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

The adaptive service model (AdaptiveService) includes cognitive information base, reasoning learning base and adaptive routing, as shown in Figure 1. Among them: the cognitive information library is used for the acquisition and modeling of cognitive information, the inference learning library is used for the prediction and reasoning of resource allocation sub-decisions, adaptive routing combined with packet scheduling and channel allocation to make global optimal routing decisions, and control the entire Allocation of network resources. The key component of the adaptive service model is adaptive routing, which controls the allocation process of network resources by exchanging a series of cognitive information. Specifically, it includes three stages: (1) Cognitive information acquisition and modeling. Each network node adds the cognitive information of each layer to the cognitive information database and models them in a unified manner. (2) Resource allocation sub-decision prediction and reasoning. According to the cognitive information in the cognitive information base, generate corresponding packet scheduling strategies and channel allocation strategies, as well as corresponding candidate sets. (3) Global optimal routing decision generation. Based on the selected data packets and available channels in the candidate set, the next hop node is selected.

The cognitive information base includes service QoS requirement information, network node information, and available spectrum information. Among them: service QoS demand information includes transmission delay, delay jitter, throughput and packet loss rate. Different services have different weights in the network, and the weights of each demand index in each business are also different. In this way, the priority of each service can be obtained through weighting. The network node information stores the neighbor information of the node, which is an important reference information for routing selection. Available spectrum information includes spectrum allocation, interference constraints, and benefits.

The reasoning learning library includes packet scheduling strategy and spectrum allocation strategy. Among them: the packet scheduling strategy is the basis for each node to schedule data packets. It adjusts the weight of each service and each index in real time according to the current network changes, and dynamically adjusts the position of each data packet in the waiting queue; spectrum allocation The strategy predicts the next round of spectrum allocation based on the current spectrum allocation.

The adaptive routing is the controller of network resource allocation, controlling the scheduling and transmission channel of each data packet in the network node. It adaptively adjusts the packet scheduling strategy and spectrum allocation strategy according to the information of the current network node and neighbor nodes, finds the most suitable data packet for scheduling from the candidate data packet set, and finds the available channel from the predicted allocable spectrum, thus The optimal route is selected by joint packet scheduling and channel allocation.

The technical solutions of the present invention will be further described in detail below in conjunction with specific embodiments. Whole invention realization process is as follows:

1. Cognitive Information Acquisition and Modeling

(1) Service QoS

Service QoS requirement information includes transmission delay, delay jitter, throughput and packet loss rate to describe. Through the comprehensive judgment of these factors affecting QoS, the priority of each service is given.

业务类型的分类及权重

Classification and weight of business types

A={a ₁ ,a ₂ ,...,a _n }, a _i (i=1,2,3,...,n) represent different services, such as voice, video, data and so on.

T={t ₁ ,t ₂ ,...,t _n }, t _i (i=1,2,3,...,n) represents the weights of different services in the network.

Judgment index and weight of QoS level

C={c ₁ ,c ₂ ,...,c _n } respectively represent QoS decision indicators: transmission delay, delay jitter, throughput, packet loss rate. Each service has different requirements for QoS indicators, so the weights of each indicator are different for each service. Let the weight coefficients of each judgment index in each business be expressed as W _i =[w _i1 , w _i2 , w _i3 , w _i4 ] respectively represent the transmission delay, delay jitter, throughput, and packet loss rate of the i-th business The weight factor in .

Each business priority

Weighting is performed according to the weight of each service and the weight of each QoS index, and the priority S _i of each service is obtained as:

${\mathrm{<span\; class="notranslate">\; S</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; t</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; (</span>\underset{\mathrm{<span\; class="notranslate">\; j</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\overset{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}}{\mathrm{<span\; class="notranslate">\; c</span>}}_{\mathrm{<span\; class="notranslate">\; i</span>}}{\mathrm{<span\; class="notranslate">\; w</span>}}_{\mathrm{<span\; class="notranslate">\; ij</span>}}<span\; class="notranslate">\; )</span>$

(2) Network node information

Use the neighbor node list to store the relevant information from the nodes adjacent to this node, as shown in the network node list below:

NIDs NChannel NW Length N Postition VelocityVector ExpireTime

NID indicates the address of the neighbor node; NChannel indicates the working frequency band of the neighbor node; NWLength indicates the queue length of the packets waiting to be processed in the neighbor node; NPositinon records the geographical location of the neighbor node; VelocityVector indicates the velocity vector of the neighbor node; ExpireTime indicates the survival of the neighbor node time.

(3) Available Spectrum

Assuming that there are N cognitive users and M available spectrums, the available spectrum information in the cognitive database is as follows:

可用频谱矩阵

Available Spectrum Matrix

L={l _n,m |l _n,m ∈{0,1}} _N×M indicates the allocation of spectrum to cognitive users, l _n,m = 1 indicates that cognitive user n can use channel m

效益矩阵

benefit matrix

B={b _n,m |b _n,m ≥0} _N×M , which represents the benefit obtained by cognitive user n on channel m.

冲突矩阵

conflict matrix

C={c _n,m |c _n,m ≥0} _N×M , indicating the number of users that collide with cognitive user n on channel m.

干扰约束矩阵

interference constraint matrix

I={i _n,k,m |i _n,k,m ∈{0,1}} _N×N×M , which means that for a certain available channel, if different cognitive users use it at the same time, the Interference may occur. i _n,k,m =1 indicates that interference will be generated.

2. Prediction and reasoning of resource allocation sub-decisions

(1) Group scheduling strategy

When considering the grouping scheduling strategy, the waiting time of the current grouping must also be considered, so as to prevent some low-priority services from being starved because they cannot be scheduled. Thus, the scheduling priority of each service group is obtained as Q _i =cS _i +(1-c)R _i , where c is the adjustment coefficient and R _i is the waiting time of the group. The implementation process of group scheduling strategy is as follows:

The first step is to update the waiting time of all packets on the node.

In the second step, the priorities Q _i =cS _i +(1-c)R _i of all packets on the node are updated.

The third step is to adjust the position of each packet on the waiting queue according to the latest packet priority, and sort according to the priority from high to low.

The fourth step is to select the N groups with the highest priority as candidate groups.

(2) Spectrum strategy

The specific implementation process is as follows:

The first step is to create an available spectrum matrix L, a benefit matrix B, a conflict matrix C, and an interference constraint matrix I.

In the second step, a non-interference allocation matrix T is created according to the available spectrum matrix L, the conflict matrix C, and the interference constraint matrix I.

The third step is to create an optimized allocation matrix R according to the non-interference allocation matrix T and the benefit matrix B.

The fourth step is to allocate channels according to business priorities according to the optimized allocation matrix R.

In the fifth step, the available spectrum matrix L, the conflict matrix C, the interference constraint matrix I, and the optimized allocation matrix R are updated for the next round of spectrum allocation.

3. Global Optimal Routing Decision Generation

Adaptive routing is the core of the adaptive service model. When receiving a business request from the application layer, it drives the packet scheduling strategy and spectrum strategy of the reasoning learning library in parallel, and then makes a global decision based on the candidate sub-decisions returned by the reasoning learning library optimal routing decisions. The specific implementation process is as follows:

In the first step, the adaptive routing waits for the service request from the application layer. When there is packet data to be transmitted, the data packet and the available spectrum are selected from the corresponding candidate scheduling packet set and the available spectrum set; during this period, if any candidate set If it is empty, go to the second step;

In the second step, the corresponding sub-decision is called to generate a candidate set in the current state, and the generated candidate set is returned to the adaptive routing. Adaptive routing selects the next hop node for the data packet according to the current cognitive information and the candidate set.

The selection of the next hop node has the following principles: (1) the shortest distance, (2) the most stable link, (3) the shortest waiting queue length, and (4) the most stable channel. The distance between the current node and the neighbor nodes is represented _by d _i ; if the current node and the neighbor nodes are close to each other, it indicates that the link is relatively stable; The waiting queue length of the data packet in the neighbor node is expressed by wl _i , according to the formula, the priority selection weight of the neighbor node can be expressed as: p _i =c ₁ d _i +c ₂ rd _i +(1-c ₁ -c ₂ )wl _i , c ₁ , c ₂ are the adjustment coefficients, and the neighbor node with the smallest priority value and the channel between the current node and the optimal allocation matrix R will be the next hop node.

Specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art may make various changes or modifications within the scope of the claims, which do not affect the essence of the present invention.

Claims (4)

1. A joint grouping scheduling and channel assignment routing method based on an adaptive service model, characterized in that, comprising sequentially connected cognitive information base, reasoning learning base and adaptive routing scheduler, wherein: Cognitive information base is used to acquire and model cognitive information; The reasoning learning library is used for prediction and reasoning of resource allocation sub-decisions; The adaptive routing scheduler is used for joint packet scheduling and channel allocation to make global optimal decisions and control the allocation of the entire network resources. 2. The joint packet scheduling and channel allocation routing method based on the adaptive service model according to claim 1, wherein the cognitive information library includes service QoS requirement information, network node information, and available spectrum information, wherein: Service QoS requirement information includes transmission delay, delay jitter, throughput and packet loss rate; The network node information includes the neighbor information of the node; Available spectrum information includes spectrum allocation, interference constraints, and benefits. 3. The joint packet scheduling and channel allocation routing method based on the adaptive service model according to claim 1, wherein the inference learning library includes a packet scheduling strategy module and a spectrum allocation strategy module, wherein: The packet scheduling strategy module is used to adjust the weight of each service and each index in real time according to the current network changes, and dynamically adjust the position of each data packet in the waiting queue; The spectrum allocation strategy module is used to predict the next round of spectrum allocation according to the current spectrum allocation situation. 4. The joint packet scheduling and channel allocation routing method based on the adaptive service model according to claim 1, characterized in that, the adaptive routing scheduler is the controller of network resource allocation, and controls the routing of each data packet in the network. Node scheduling, routing, and transmission channels, the adaptive routing scheduler adaptively adjusts the packet scheduling strategy and spectrum allocation strategy according to the information of the current network node and neighbor nodes, and finds the most suitable data for scheduling from the candidate data packet set Grouping, finding available channels from the predicted allocatable spectrum, so that the joint packet scheduling and channel allocation will transmit the data packets of the current node from the optimal route.

Images