Download presentation

Presentation is loading. Please wait.

Published byJamison Whiteaker Modified over 2 years ago

1
A Centralized Scheduling Algorithm based on Multi-path Routing in WiMax Mesh Network Yang Cao, Zhimin Liu and Yi Yang International Conference on Wireless Communications, Networking and Mobile Computing, 2006. WiCOM 2006. Mei-jhen Chen

2
Outline Introduction Scheduling in WiMax Mesh Network Centralized Scheduling based on Multi-path Routing –Basic Ideas –Multi-path Routing –Centralized Scheduling Performance Analysis Conclusions

3
Introduction The IEEE 802.16 standard is one of the most promising technologies in the future. –PMP (Point to Multipoint) mode –Mesh mode Enhancing the capability of WiMAX Mesh network is one of the focal problem nowadays. The appropriate routing and scheduling algorithm could effectively enhance the throughput of WMN (wireless mesh network).

4
Introduction (cont.) Nobody considered the route metrics and scheduling schemes together, as well as the network load balance and QoS. Authors introduce the cross-layer concept. An effective centralized scheduling algorithm based on multi-path routing (MR-CS) in WiMAX mesh network was proposed. The network load balance, spatial reuse and QoS guarantee are synthetically considered.

5
Scheduling in WiMax Mesh Network -control messages Mesh BSMesh SSs MSH-CSCH (Request message) MSH-CSCH (Grant message) Mesh BS collects the bandwidth requests, determines the resource allocation for each link The SSs determine its actual uplink and downlink transmission time Mesh SS estimates its resource request and sends it MSH-CSCF (broadcast)

6
Centralized Scheduling based on Multi-path Routing -Basic Ideas In the centralized scheduling –Mesh BS Determine the optimized network routing tree and resource scheduling scheme. –Mesh SS receive the corresponding scheduling message behave based on it strictly. Designing the routing and scheduling scheme of the Mesh BS is the core problem of the algorithm.

7
Centralized Scheduling based on Multi-path Routing -Basic Ideas Assumptions –Nodes can not send and receive data at the same time. –Nodes can not send or receive data in the signal range of communicating nodes because of interference. –Non-interference links can communicate concurrently. –The topology doesn’t make any change during the scheduling period. –The control and scheduling sub-frame are long enough. –The traffic is always between the Mesh SS and Mesh BS.

8
Centralized Scheduling based on Multi-path Routing -Basic Ideas The basic idea of MR-CS is the cross-layer design between the network (routing) and MAC (scheduling) layers.

9
Centralized Scheduling based on Multi-path Routing -Multi-path Routing The Multi-path Routing Module includes two steps: –The available routes search –The optimized route selection

10
Centralized Scheduling based on Multi-path Routing -Multi-path Routing - The available routes search The Multi-path Source Routing (MSR) could be adopted by the routes search process to find the available Multi-path Routes for each Mesh SS. R x ={R x (1), R x (2), …, R x (N)} : for a Mesh SS x, its Multi-path Routes –N : the number of available Multi-path Routes. I x-y : the interference table of any link connecting Mesh SS x and Mesh SS y L x-y : the link between Mesh SS x and SS y N(x) : the neighbor muster of Mesh SS x For any link L i-j –If i=x or i=y or j=x or j=y, j ∈ N(x), I ∈ N(y). L i-j ∈ I x-y

11
Centralized Scheduling based on Multi-path Routing -Multi-path Routing - The optimized route selection R x (i) –The i-th Multi-path Routes of Mesh SS x. –R x (i) = {L x-j1, L j1-j2, …, L jk-B } Interference Factor I x (i) –the interference link amount muster of every link L x-y along the route Rx(i) –I x (i) = {C(I x-j1 ), C(I j1-j2 ), …, C(I jk-B )}, –C(I x-y ) : the number of elements in I x-y. Load Factor P x (i) –the resource demand muster of every link L x-y along the route R x (i), –i.e. P x (i) = {P x-j1, P j1-j2, …, P jk-B }. –P x-y, counting in minislots including source and forward traffic QoS Factor Q x (i) –the QoS index(Q x ) muster of every node along the route R x (i). –Q x = (P x1 · 3+ P x2 · 2+ P x3 · 1) / (P x1 + P x2 + P x3 ).

12
Centralized Scheduling based on Multi-path Routing -Multi-path Routing - The optimized route selection the Routing Factor S x (i) –S x (i) = sum( Ix(i) · Px(i) · Qx(i) ) = C(I x-j1 )*P x-j1 *Q x + C(I j1-j2 )*P j1-j2 *Q j1 + … + C(I jk-B )*P jk-B *Q jk The R x (i) with the least S x (i) in Multi-path Routes R x will be chosen as the Optimized Route.

13
Centralized Scheduling based on Multi-path Routing -Centralized Scheduling To assign the minislots for every Mesh SS. –Pending Links, PL t : the link muster with data transmission demand in minislot t. –Scheduling Links, SL t : the link muster having been scheduled to transmit in minislot t.

14
Centralized Scheduling based on Multi-path Routing -Centralized Scheduling Always choose the highest QoS class as preference. In the 1 st minislot, –PL 1 = {L x1-y1 (c 1 ), L x2-y2 (c 2 ), …, L xn-yn (c n )} L xi-yi (c i ) : that the link L xi-yi demands c i minislots to transmit the traffic. –Choose the link with the largest c k, as the Preferential Link, PrL 1 = L xk-yk (c k ), and put it into PL Then find the links that can communicate concurrently. –Arrange them into the Scheduling Links, –SL 1 ={L xi-yi | L xi-yi ! ∈ IPrL 1 ∩ L xi-yi ∈ PL t, i=1, 2, …} Updating the Pending Links PL t, delete the links that has been transmitted and add the new ones because of packet forwarding Repeat the above steps in the next minislot till all the scheduling has been accomplished, i.e. PL t = Φ.

15
Performance Analysis -example BS SS1SS2SS3 SS4SS5SS6 SS7 SS8SS9SS10 SS11 SS12SS13 1 (1,0,0) 2 (1,1,0) 1 (1,0,0) 2 (1,1,0) 3 (1,1,1)

16
Performance Analysis -example

18
Performance Analysis -Simulation Matlab tool Random topology is generated in a round area with Radius –N is the number of Mesh SSs –d is the maximal transmission between two nodes Assume the single channel in the network without any bit errors, and the nodes fixed. transmission data rate : 51.2Mbps

19
Performance Analysis -Simulation There are 3 types of traffic –real-time CBR The packet length of CBR is 440bytes, with constant interval of 10ms. –real-time VBR the ON/OFF model setting 1s to the average ON time and 1.35s to the average OFF time. The VBR packet length is 160ms, coming every 20ms in the ON time. –data flows Poisson stream with rate λ according to the bandwidth. Every node in the network has the equal traffic load.

20
Performance Analysis -Simulation

23
Conclusions Authors propose a centralized scheduling algorithm based on multi-path routing (MR-CS) in WiMAX mesh network. The network load balance, spatial reuse and QoS guarantee are synthetically considered. The analysis result shows this algorithm has greatly improved the network throughput and the efficiency of the centralized scheduling.

Similar presentations

OK

Exploiting Spectral Reuse in Resource Allocation, Scheduling,and Routing for IEEE 802.16 Mesh Networks Lien-Wu Chen, Yu-Chee Tseng Department of Computer.

Exploiting Spectral Reuse in Resource Allocation, Scheduling,and Routing for IEEE 802.16 Mesh Networks Lien-Wu Chen, Yu-Chee Tseng Department of Computer.

© 2017 SlidePlayer.com Inc.

All rights reserved.

Ads by Google

File type-ppt on cybercrime Ppt on bluetooth based smart sensor networks machine Ppt on cross cultural communication barriers Ppt on role of mass media in education Ppt on working of nuclear power plant in india Ppt on cloud computing in education Ppt on business planning Ppt on contact management system in java Ppt on bmc remedy action Ppt on revolution of the earth and seasons