![ACEEE Int. J. on Network Security, Vol. 01, No. 03, Dec 2010
Performance Analysis and Development of an
Efficient Routing Scheme for IEEE
802.16/WiMAX Mesh Networks
Rijvi Ahmed; A.H.M. Nazir Hossain; M.M. Shahriar Maswood
Department of Electronics & Communication Engineering, Khulna University of Engineering & Technology
Khulna- 9203, Bangladesh. nazir_kuetcec05@yahoo.com, r_shaon@yahoo.com
Abstract— A collision-free centralized scheduling algorithm II. BACKGROUND
for IEEE 802.16 WiMAX provides a mechanism for
creations multi-hop mesh and high-quality wireless Performances of wireless mesh networks can be
multimedia services which can be developed as a high speed improved by employing spatial reuse with concurrent
broad band wireless network. In centralized scheduling for transmission. A vast amount of research has been
IEEE 802.16 mesh networks, all packets should be conducted in routing and scheduling in wireless networks.
transported through the BS (Base station).The links to or Jun wong, Weijia Jia, Liuseng Huamg et al [4] proposed an
from the BS becomes the system’s bottleneck and the
efficient centralized scheduling algorithm for IEEE 802.16
throughput is heavily impacted by the interference. To solve
this problem, we evaluated the proposed algorithm with five
mesh radio networks. They used TDMA concept. In a time
selection criteria of scheduling through extensive simulations slot, those nodes will transmit which are not interfered by
and the experimental results are instrumental for improving each others. They proposed four node selection criteria –
the performance of IEEE 802.16 based WMNs in terms of link nearest, farthest, minimum interference and random. In our
scheduling. We compared the effect of two routing and one paper, we mentioned this algorithm as R_A_1. According
scheduling algorithm on the scheduling length. The result to this routing algorithm, when a new node joins to the
shows that best algorithm has improved the system network it will select a node of minimum node id from the
performance in the aspects of scheduling length, transmission neighbors of that node. But through this, more time slots
range, and channel utilization ratio.
may be needed to transmit/receive 1 packet to/from BS. To
Keywords- IEEE 802.16, Mesh Network, WiMAX, Multi-hop solve this problem, Hung-Yu, Samrat Ganguly and Rauf
Routing, Mesh mode. Izmailov Wei proposed maximum weighted nodeselection
and minimum blocking metrics routing algorithm in their
scheduling algorithm. We mentioned this routing algorithm
I. INTRODUCTION as R_A_2 in this paper later. But, some problem remains
The IEEE 802.16 standard, commonly known as still like ambiguity problem & Imperfect route selection.
WiMAX, was published to construct the last-mile wireless We have proposed our algorithm to solve this type of
broadband access (WBA) in metropolitan area networks problems which is indicated as R_A_3.
and provides better performance comparable to other
traditional cable, DSL, or T1networks. In this paper, we III. PROPOSED ALGORITHM
consider the problem of routing and scheduling packets We proposed a routing algorithm which removes the
based on centralized scheme. In WiMAX mesh centralized problems of R_A_2. The ambiguity problem is removed by
routing and scheduling, the BS determines the routes of all two steps. If there are more than one neighbor nodes with
SSs and also decides the transmission sequence in data minimum blocking metrics, then select a node among them
subframes. The scheduling algorithm, along with the policy which have minimum hops to/from BS & if there are also
of routing tree construction decides the system more than one node with minimum hops no from/to BS,
performance. System performance is evaluated by the then select the node of minimum node id among them. To
parameter scheduling length (where lower scheduling remove imperfect node selection problem it is needed to
length indicates higher performance). So the problem is update all previous node’s route when a new node is
that how to reduce the scheduling length by designing inserted. When a new node joins in the networks, then all
efficient routing and scheduling protocol. In centralized the nodes are arranged in a list in ascending order to their
operation we find no scope in scheduling. Therefore the distance from BS and all nodes will be set as
problem is to develop an efficient routing algorithm for unselected/unrouted nodes. Each node will be scanned
centralized mesh mode of operation. In our simulation we from list and their route (i.e. parent) will be selected
consider only uplink transmission and it can be easily according to blocking metric concept (as R_A_2). That
converted to downlink transmission. In simulation we don’t means, the node will find the selected nodes from its
consider the condition of link failure; because our proposed neighbors and select one of them who has minimum
algorithm is not aware of link failure. Moreover this blocking metric as its parent
routing algorithm is proposed only for fixed nodes. So R_A_3
there is scope to work on routing of centralized mesh 1. Set BS as selected/routed node
network for mobile nodes with considering link failure. Do step 2 to step 4 for each new node when it is inserted:
© 2010 ACEEE 12
DOI: 01.IJNS.01.03.50](https://image.slidesharecdn.com/50-120925230252-phpapp02/85/Performance-Analysis-and-Development-of-an-Efficient-Routing-Scheme-for-IEEE-802-16-WiMAX-Mesh-Networks-1-320.jpg)
![ACEEE Int. J. on Network Security, Vol. 01, No. 03, Dec 2010
REFERENCES in IEEE 802.16 mesh mode”, in: Proc. of the Sixth ACM
International Symposium on Mobile Ad Hoc Networking and
[1] P. Kyasanur, N.H. Vaidya, “Routing and interface assignment Computing (MOBIHOC 2005), May 2005, pp. 78–89.
in multi-channel multi-interface wireless networks”, in: Proc. of [5] H. Shetiya and V. Sharma, “Algorithms for routing and
the 2005 IEEE Wireless Communications and Networking centralized scheduling to provide qos in ieee 802.16 mesh
Conference (WCNC 2005), March 2005, pp. 2051–2056. networks,” in WMuNeP’05, 2005, pp. 140–149.
[2] A. Raniwala, T.-C. Chiueh, “Architecture and algorithms for [6] H. Wei, S. Ganguly, R. Izmailov, and Z. J. Hass,
an IEEE 802.11-based multi-channel wireless mesh network”, in: “Interference-aware ieee 802.16 wimax mesh networks,” in
Proc. of the 24th Annual Joint Conference of the IEEE Computer VTC’05, 2005, pp. 3102–3106.
and Communications Societies (INFOCOM 2005), March 2005, [7] D. Aguayo, J. Bicket, S. Biswas, G. Judd, R. Morris, “Link-
pp. 2223–2234. level measurements from an 802.11b mesh network”, in: Proc. of
[3] V. Gambiroza, B. Sadeghi, E. Knightly, “End-to-end the 2004 ACM Annual Conference of the Special Interest Group
performance and fairness in multihop wireless backhaul on Data Communication (SIGCOMM), August 2004, pp. 121–
networks”, in: Proc. of the 10th ACM Annual International 132.
Conference on Mobile Computing and Networking (MOBICOM [8] H. Lim, C. Lim, and J. C. Hou, “A coordinate-based
2004), Sep. 2004, pp. 287–301. approach for exploiting temporal-spatial diversity in wireless
[4] M. Cao, W.C. Ma, Q. Zhang, X.D. Wang, W.W. Zhu, mesh networks,” in MobiCom’06, 2006, pp. 14–25.
“Modeling and performance analysis of the distributed scheduler
© 2010 ACEEE 14
DOI: 01.IJNS.01.03.50](https://image.slidesharecdn.com/50-120925230252-phpapp02/85/Performance-Analysis-and-Development-of-an-Efficient-Routing-Scheme-for-IEEE-802-16-WiMAX-Mesh-Networks-3-320.jpg)
Performance Analysis and Development of an Efficient Routing Scheme for IEEE 802.16/WiMAX Mesh Networks
- 1. ACEEE Int. J. on Network Security, Vol. 01, No. 03, Dec 2010 Performance Analysis and Development of an Efficient Routing Scheme for IEEE 802.16/WiMAX Mesh Networks Rijvi Ahmed; A.H.M. Nazir Hossain; M.M. Shahriar Maswood Department of Electronics & Communication Engineering, Khulna University of Engineering & Technology Khulna- 9203, Bangladesh. [email protected], [email protected] Abstract— A collision-free centralized scheduling algorithm II. BACKGROUND for IEEE 802.16 WiMAX provides a mechanism for creations multi-hop mesh and high-quality wireless Performances of wireless mesh networks can be multimedia services which can be developed as a high speed improved by employing spatial reuse with concurrent broad band wireless network. In centralized scheduling for transmission. A vast amount of research has been IEEE 802.16 mesh networks, all packets should be conducted in routing and scheduling in wireless networks. transported through the BS (Base station).The links to or Jun wong, Weijia Jia, Liuseng Huamg et al [4] proposed an from the BS becomes the system’s bottleneck and the efficient centralized scheduling algorithm for IEEE 802.16 throughput is heavily impacted by the interference. To solve this problem, we evaluated the proposed algorithm with five mesh radio networks. They used TDMA concept. In a time selection criteria of scheduling through extensive simulations slot, those nodes will transmit which are not interfered by and the experimental results are instrumental for improving each others. They proposed four node selection criteria – the performance of IEEE 802.16 based WMNs in terms of link nearest, farthest, minimum interference and random. In our scheduling. We compared the effect of two routing and one paper, we mentioned this algorithm as R_A_1. According scheduling algorithm on the scheduling length. The result to this routing algorithm, when a new node joins to the shows that best algorithm has improved the system network it will select a node of minimum node id from the performance in the aspects of scheduling length, transmission neighbors of that node. But through this, more time slots range, and channel utilization ratio. may be needed to transmit/receive 1 packet to/from BS. To Keywords- IEEE 802.16, Mesh Network, WiMAX, Multi-hop solve this problem, Hung-Yu, Samrat Ganguly and Rauf Routing, Mesh mode. Izmailov Wei proposed maximum weighted nodeselection and minimum blocking metrics routing algorithm in their scheduling algorithm. We mentioned this routing algorithm I. INTRODUCTION as R_A_2 in this paper later. But, some problem remains The IEEE 802.16 standard, commonly known as still like ambiguity problem & Imperfect route selection. WiMAX, was published to construct the last-mile wireless We have proposed our algorithm to solve this type of broadband access (WBA) in metropolitan area networks problems which is indicated as R_A_3. and provides better performance comparable to other traditional cable, DSL, or T1networks. In this paper, we III. PROPOSED ALGORITHM consider the problem of routing and scheduling packets We proposed a routing algorithm which removes the based on centralized scheme. In WiMAX mesh centralized problems of R_A_2. The ambiguity problem is removed by routing and scheduling, the BS determines the routes of all two steps. If there are more than one neighbor nodes with SSs and also decides the transmission sequence in data minimum blocking metrics, then select a node among them subframes. The scheduling algorithm, along with the policy which have minimum hops to/from BS & if there are also of routing tree construction decides the system more than one node with minimum hops no from/to BS, performance. System performance is evaluated by the then select the node of minimum node id among them. To parameter scheduling length (where lower scheduling remove imperfect node selection problem it is needed to length indicates higher performance). So the problem is update all previous node’s route when a new node is that how to reduce the scheduling length by designing inserted. When a new node joins in the networks, then all efficient routing and scheduling protocol. In centralized the nodes are arranged in a list in ascending order to their operation we find no scope in scheduling. Therefore the distance from BS and all nodes will be set as problem is to develop an efficient routing algorithm for unselected/unrouted nodes. Each node will be scanned centralized mesh mode of operation. In our simulation we from list and their route (i.e. parent) will be selected consider only uplink transmission and it can be easily according to blocking metric concept (as R_A_2). That converted to downlink transmission. In simulation we don’t means, the node will find the selected nodes from its consider the condition of link failure; because our proposed neighbors and select one of them who has minimum algorithm is not aware of link failure. Moreover this blocking metric as its parent routing algorithm is proposed only for fixed nodes. So R_A_3 there is scope to work on routing of centralized mesh 1. Set BS as selected/routed node network for mobile nodes with considering link failure. Do step 2 to step 4 for each new node when it is inserted: © 2010 ACEEE 12 DOI: 01.IJNS.01.03.50
- 2. ACEEE Int. J. on Network Security, Vol. 01, No. 03, Dec 2010 2. Sorting all nodes in ascending order according to transmission range of all nodes to 7 units & minimum distance from BS. separation between nodes 3 unit with variable node no 3. Reset all nodes as unselected/unrouted node. whereas in fig-2, we have varied transmission range of the 4. Scanning SSs in ascending order of distance from BS nodes in a particular area. From fig.1, it is clear that no. of and find route of all SSs like below (do step ‘a’ to ‘f’ for scheduling length increases more for R_A_3 than R_A_1 each SS): & R_A_2 with the increment of the no. of nodes a. Find neighbors and then selected nodes from . R_A_1 neighbors. R_A_2 b. Find the node with minimum blocking metrics from Scheduling length 120 R_A_3 selected nodes. c. Select the node with minimum hop, if there is more 80 than one node in step 'b'. d. Select a node with minimum id, if there is more than one node in step 'c'. 40 40 60 80 100 120 e. Set the node in step’d’ as parents of scanned SS. Nodes number f. Set the scanned SS as routed/selected node. For R_A_2, route time for n no. of node is nt, where t is Fig. 1. Comparison among R_A_1, R_A_2 & R_A_3 in the Fig-1: . average route calculation time, when a new node is Comparison among R_A_1, R_A_2 & R_A_3 in the aspect of scheduling inserted. But, for R_A_3, Route calculation time of i-th length varying node no node = s + i × t ≅ i × t ; For n nodes, total route calculation R_A_1 time =(s +t) +(s +2 ) +(s +3 ) +..........(s +(n−1t) +(s +nt t t + ) ) R_A_2 Scheduling length 52.4 R_A_3 n × ( n + 1) = n×s + ×t . 2 50.7 Route time for n - th node (R_A_3) n×t ∴ = =n Route time for n - th node (R_A_2) t 49 n × (n + 1) 4 8 12 16 n×s + ×t Transmission range Total route time of n nodes (R_A_3) 2 = Total route time of n nodes (R_A_2) n∗t Fig. 2. Comparison among R_A_1, R_A_2 & R_A_3 in the aspect of n +1 scheduling length varying transmission range ≅ 2 From fig-2, it is clear that when transmission range Here route calculation time of n-th node using R_A_3 is increases then scheduling length for all routing algorithm n times than R_A_2. So it may be seemed that, calculation becomes closer and for a large value of transmission range time is a problem for R_A_3, especially when node number it becomes same. This is because when the transmission (n) is very large. But interestingly notice that, total route range is too large then all the nodes stay in the transmission calculation time of n nodes using R_A_3 is (n+1)/2 times range of BS and hence the network operates like PMP than R_A_2. Therefore the ratio of calculation time has (point to multi point mode) mode. Then the scheduling significantly decreased for total route calculation. length becomes equal to nodes number, because in PMP Moreover routing algorithm does not run all times; rather it modes at each time slot only one node directly transmit only runs when a new node joins or when changes occur in to/from BS. the network topology. So, more calculation time is not a major problem. V. SUMMARY IV. PERFORMANCE EVALUATION We proposed a collision-free centralized scheduling algorithm for IEEE 802.16 based WMNs. This scheduling Our simulation is based on centralized scheduling and scheme takes fairness, channel utilization and transmission one directional transmission. The length of scheduling is delay into consideration. In the proposed algorithm, the the most important performance measure of a selection policy for scheduled links will impact the scheduling/routing algorithm and it is considered in most of algorithm’s performance. We use the length of scheduling, the existing literatures. From our simulation first we find and transmission range to evaluate the performance of the which scheduling criterion is best. We have developed a proposed scheduling algorithm. Our future work will simulation model using C programming language. Using mainly focus on the problem of link failure and dynamic this model any routing or scheduling algorithm can be run route construction. on a network topology and gives the output in terms of scheduling metric, CUR, scheduling length. It gives the visualized output of the network. Finally we will show that how network performance vary (using R_A_3, R_A_2 & R_A_1) according to no. of nodes & transmission range between nodes. In fig-1, we have considered the © 2010 ACEEE 13 DOI: 01.IJNS.01.03.50
- 3. ACEEE Int. J. on Network Security, Vol. 01, No. 03, Dec 2010 REFERENCES in IEEE 802.16 mesh mode”, in: Proc. of the Sixth ACM International Symposium on Mobile Ad Hoc Networking and [1] P. Kyasanur, N.H. Vaidya, “Routing and interface assignment Computing (MOBIHOC 2005), May 2005, pp. 78–89. in multi-channel multi-interface wireless networks”, in: Proc. of [5] H. Shetiya and V. Sharma, “Algorithms for routing and the 2005 IEEE Wireless Communications and Networking centralized scheduling to provide qos in ieee 802.16 mesh Conference (WCNC 2005), March 2005, pp. 2051–2056. networks,” in WMuNeP’05, 2005, pp. 140–149. [2] A. Raniwala, T.-C. Chiueh, “Architecture and algorithms for [6] H. Wei, S. Ganguly, R. Izmailov, and Z. J. Hass, an IEEE 802.11-based multi-channel wireless mesh network”, in: “Interference-aware ieee 802.16 wimax mesh networks,” in Proc. of the 24th Annual Joint Conference of the IEEE Computer VTC’05, 2005, pp. 3102–3106. and Communications Societies (INFOCOM 2005), March 2005, [7] D. Aguayo, J. Bicket, S. Biswas, G. Judd, R. Morris, “Link- pp. 2223–2234. level measurements from an 802.11b mesh network”, in: Proc. of [3] V. Gambiroza, B. Sadeghi, E. Knightly, “End-to-end the 2004 ACM Annual Conference of the Special Interest Group performance and fairness in multihop wireless backhaul on Data Communication (SIGCOMM), August 2004, pp. 121– networks”, in: Proc. of the 10th ACM Annual International 132. Conference on Mobile Computing and Networking (MOBICOM [8] H. Lim, C. Lim, and J. C. Hou, “A coordinate-based 2004), Sep. 2004, pp. 287–301. approach for exploiting temporal-spatial diversity in wireless [4] M. Cao, W.C. Ma, Q. Zhang, X.D. Wang, W.W. Zhu, mesh networks,” in MobiCom’06, 2006, pp. 14–25. “Modeling and performance analysis of the distributed scheduler © 2010 ACEEE 14 DOI: 01.IJNS.01.03.50