Routing Metrics and Protocols for Wireless Mesh Network

1. IEEE Network • January/February 2008 Routing Metrics and Protocols for Wireless Mesh Network Miguel Elias M. Campista Pedro Miguel Esposito Igor M. Moraes Luís Henrique M. K. Costa Otto Carlos M. B. Duarte Diego G. Passos Célio Vinicius N. de Albuquerque Débora Christina M. Saade Marcelo G. Rubinstein Instructor : Professor Frank Y.S. Lin Presented by Ray J.P. Lo NTU IM OPLab

2. Agenda • Authors • Introduction • Routing Metrics • Routing Protocols • Mesh Network Performance Analysis • Conclusion NTU IM OPLab

3. Authors NTU IM OPLab

4. Authors’ Biographies • Universidade Federal do Rio de Janeiro • Miguel Elias M. Campista • Pedro Miguel Esposito • Igor M. Moraes • Luís Henrique M. K. Costa • Otto Carlos M. B. Duarte • Universidade Federal Fluminense • Diego G. Passos • Célio Vinicius N. de Albuquerque • Débora Christina M. Saade • Universidade do Estado do Rio de Janeiro • Marcelo G. Rubinstein NTU IM OPLab

5. Introduction NTU IM OPLab

6. Why Wireless Networks ? • Wireless networks are becoming increasingly popular as they provide flexibility, mobility support, and are easy to deploy. • In addition, the reduced wired infrastructure combined with large-scale commercialization results in plummeting costs. • More and more ISPs offer wireless access that in the long term will result in ubiquitous Internet. NTU IM OPLab

7. Wireless Networks • Infrastructure-based wireless networks • ex: IEEE 802.11 wireless distribution systems. • Limit the coverage to users within the transmission range of access points. • Access points are connected to a wired network, which incurs high infrastructure costs. • Ad hoc networks • Have no infrastructure costs. • Nevertheless, ad hoc networks cannot supply backhaul access and may become a collection of isolated networks due to user mobility. NTU IM OPLab

8. Wireless Mesh Networks (WMNs) • WMNs aim at guaranteeing connectivity. • WMNs build a multihop wireless backbone to interconnect isolated LANs and to extend backhaul access to users not within range of typical access points. • Backbone routers are usually stationary, so they can be permanently power-supplied and permit routing metrics to model link quality. NTU IM OPLab

9. Routing Metrics NTU IM OPLab

10. Hop Count • Ad hoc networks usually use the hop count as a routing metric. • It is appropriate for ad hoc networks because new paths must be found rapidly. • This is important in ad hoc networks because of user mobility. NTU IM OPLab

11. WMN Routing Metrics • ETX • ML • ETT • WCETT • mETX • ENT • iAWARE NTU IM OPLab

12. ETX(expected transmission count) • ETX is the first metric proposed for WMNs. • ETX is the expected number of transmissions a node requires to successfully transmit a packet to a neighbor. • Each node periodically broadcasts probes containing the number of received probes from each neighbor. • The number of received probes is calculated at the last Ttime interval in a sliding-window fashion. NTU IM OPLab

13. ETX(expected transmission count) • Example: The ETX of link AB • Considering the delivery ratio of probes sent on the forward (df) and reverse (dr) directions. • Delivery ratios are, at the same T interval, • The fraction of successfully received probes from A announced by B. • The fraction of successfully received probes from B. • The ETX of link ABis 1/(df× dr). NTU IM OPLab

14. ETX(expected transmission count) • Assume that : • Each node broadcasts a probe packet every second. • Node A has received 8 probe packets from B in the previous 10 seconds. • In the last probe packet, B reported that it had received 9 probe packets from A in the previous 10 seconds. • Thus, ETX = 1/(0.9 * 0.8) = 1.39. => dr = 0.8 => df = 0.9 NTU IM OPLab

15. ETX(expected transmission count) • The ETX computation considers both forward and reverse directions because of data- and ACK-frame transmission. • The chosen route is the one with the lowest sum of ETX along the route to the destination. NTU IM OPLab

16. ML(minimum loss) • The ML metric also is based on probing to compute the delivery ratio. • Rather than calculating ETX, ML finds the route with the lowest end-to-end loss probability. • ML multiplies the delivery ratios of the links in the reverse and forward directions to find the best path. • The use of multiplication reduces the number of route changes, improving network performance. NTU IM OPLab

17. The Shortcomings of ETX • Because broadcasts usually are performed at the network basic rate, and probes are smaller than typical data packets, the implementation of ETX has revealed two shortcomings: • It doesn’t distinguish links with different bandwidths. • It doesn’t consider data-packet sizes. • Thus, the network is operating at low rates, the performance of ETX becomes low. NTU IM OPLab

18. ETT (expected transmission time) • To cope with the issues caused by ETX, the expected transmission time (ETT) was proposed. • ETT is the time a data packet requires to be transmitted successfully to each neighbor. NTU IM OPLab

19. ETT (expected transmission time) • Two main approaches to compute ETT: • The approach proposed in “Routing in Multi-Radio, Multi-Hop Wireless Mesh Networks” • ETT is the product between ETX and the average time a single data packet requires to be delivered. • The approach proposed in “High-Throughput Routing for Multi-Hop Wireless Networks” • It estimates the loss probability by considering that IEEE 802.11 uses data and ACK frames. NTU IM OPLab

20. The 1st Approach to Compute ETT • ETT= ETX × t • t = S / B • S = A fixed data-packet size. • B = The estimated bandwidth of each link. • The authors prefer to periodically estimate the bandwidth (B) than to use rates retrieved from firmware. NTU IM OPLab

21. The Packet-Pair Technique Used to Calculate B • Each node sends a sequence of two back-to-back packets (a small one (137 bytes) followed by a large one (1,137bytes) ) to each of it’s neighbors every minute. • Each neighbor measures the inter-arrival period between the two packets and reports it back to the sender. • The sender takes the minimum of 10 consecutive samples. • B = The size of the large packet of the sequence / The minimum delay received for that link. NTU IM OPLab

22. The 2nd Approach to Compute ETT • The idea is to periodically compute the loss rate of data and ACK frames to each neighbor. • Data frames : estimated by broadcasting a number of packets of the same size as data frames, one packet for each data rate defined in IEEE 802.11. • ACK frames : estimated by broadcasting small packets of the same size as ACK frames and sent at the basic rate that is used for ACKs. NTU IM OPLab

23. The 2nd Approach to Compute ETT • ETT =1/ ( P(ack) * rt ) , rt = max(r1, r2, r5.5, r11) • In Total, it sends each of the below five probes at independent random intervals averaging ten seconds : • one 32-byte probe at 1 megabit • one 1500-byte probe at each of 1, 2, 5.5, and 11 megabits • P(ack) is the delivery probability of an ACK based on probe losses in the reverse direction. • rt is the best throughput of broadcast packets in the forward direction at bit-rate t megabits/second. • Similarly to ETX, the chosen route is the one with the lowest sum of ETT values. NTU IM OPLab

24. The Use of Multiple Channels in WMNs • It is possible to improve network throughput by using, at the same time, the available non-overlapping channels defined by IEEE 802.11. • This technique must deal with two issues to become effective : • intra-flow interference • It occurs when different nodes transmitting packets from the same flow interfere with each other. • inter-flow interference • It is the interference suffered among concurrent flows. NTU IM OPLab

25. WCETT(The Weighted Cumulative ETT) • WCETT changes ETT to also consider intra-flow interference. • This metric computes end-to-end values because it must consider all channels used along the route to avoid intra-flow interference. • Its outcome is a sum of end-to-end delay and channel diversity, the final cost of the route. NTU IM OPLab

26. WCETT(The Weighted Cumulative ETT) • Consider an n-hop path, and the system has a total of k channels. Define Xjas : • Thus, Xjis the sum of transmission times of hops on channel j. NTU IM OPLab

27. WCETT(The Weighted Cumulative ETT) • We can view the above equation as a tradeoff between delay and throughput. • The first term can be considered as a measure of the latency of this path. • The second term represents the impact of bottleneck hops and can be viewed as total path throughput. • The weighted average is an attempt to strike a balance between the two. NTU IM OPLab

28. WCETT(The Weighted Cumulative ETT) • Nevertheless, WCETT neither guarantees shortest paths nor avoids inter-flow interference which may lead to choosing routes in congested areas. • The metric of interference and channel-switching (MIC) addresses these issues. NTU IM OPLab

29. MIC(The Metric of Interference and Channel-switching) • Each node takes into account the number of interfering nodes in the neighborhood to estimate inter-flow interference. • MIC also calculates its value based on the ETT metric. NTU IM OPLab

30. MIC(The Metric of Interference and Channel-switching) • N is the total number of nodes in the network. • min(ETT) is the smallest ETT in the network, which can be estimated based on the lowest transmission rate of the wireless cards. • The two components of MIC : • IRU (Interference-aware Resource Usage) • CSC (Channel Switching Cost) NTU IM OPLab

31. MIC(The Metric of Interference and Channel-switching) • Nlis the set of neighbors that the transmission on link l interferes with. • The physical meaning of IRUl is the aggregated channel time of neighboring nodes that transmissions on link l consumes. • It captures the inter-flow interference since it favors a path that consumes less channel times at its neighboring nodes. NTU IM OPLab

32. MIC(The Metric of Interference and Channel-switching) • CH(i) represents the channel assigned for node i's transmission. • prev(i) represents the channel used by the previous hop of node i along the path p. • CSC represents the intra-flow interference, since it gives paths with consecutive links using the same channel higher weights than paths that alternate their channel assignments. NTU IM OPLab

33. Fast Link-Quality Variation in WMNs • The fast link-quality variation is a critical problem of wireless networks. • Metrics based on average values computed on a time-window interval, such as ETX, may not be suitable. • This problem is more difficult in indoor environments. • To cope with this, modified ETX (mETX) and effective number of transmissions (ENT) were proposed. • These metrics consider the standard deviation in addition to link-quality average values to project physical-layer variations onto routing metrics. NTU IM OPLab

34. mETX (modified ETX) • The mETX metric also is calculated by broadcasting probes. • The difference between mETX and ETX: • ETX considers probe losses. • mETX works at the bit level. • mETX computes the bit error probability using • The position of the corrupted bit in the probe • The dependence of these bit errors throughout successive transmissions. • It’s possible because probes are composed by a previously known sequence of bits. NTU IM OPLab

35. ENT(Effective Number of Transmissions) • ENT measures the number of successive retransmissions per link considering the variance. • ENT also broadcasts probes and limits route computation to links that show an acceptable number of retransmissions according to upper-layer requirements. NTU IM OPLab

36. iAWARE(Interference Aware) • iAWARE also considers link-quality variation. • This metric uses • Signal to noise ratio (SNR) • Signal to interference and noise ratio (SINR) to continuously reproduce neighboring interference variations onto routing metrics. • It estimates the average time the medium is busy because of transmissions from each interfering neighbor. • The higher the interference, the higher the iAWARE value. • iAWARE considers intra- and inter-flow interference, medium instability, and data-transmission time. NTU IM OPLab

37. Conclusion • Although there is an increasing number of routing metrics, a consensus has not been achieved. • Up to now, most routing protocol implementations prefer metrics with simpler designs such as ETX or ETT. NTU IM OPLab

38. Routing Protocols NTU IM OPLab

39. Ad hoc Routing Protocols • Proactive • It operates like classic routing on wired networks. • Routers keep at least one route to any destination in the network. • Reactive • It request a route to a destination only when a node has a data packet to send. • Hybrid NTU IM OPLab

40. WMN Routing Protocols • Many WMN routing protocols use similar strategies used in Ad hoc routing protocols. • Most WMN routing protocols consider that the network is only composed by wireless backbone nodes. • The authors proposed a taxonomy for WMN routing protocols with four classes. • Each class mainly differs on route discovery and maintenance procedures. NTU IM OPLab

41. A Taxonomy for WMN Routing Protocols with Four Classes • Ad hoc-based • Controlled-flooding • Traffic-aware • Opportunistic NTU IM OPLab

42. Ad hoc-based Protocols • WMN ad hoc-based protocols adapt ad hoc routing protocols to deal with link-quality variations. • Routers continuously update their outgoing-link metrics and disseminate them to other routers. NTU IM OPLab

43. LQSR(Link Quality Source Routing) • LQSR is a source-routedlink-state protocol. • A link-state protocol consists of four components: • A component that discovers the neighbors of a node. • A component that assigns weights to the links a node has with its neighbors. • A component that propagates this information to other nodes in the network. • A component that uses the link weights to find a good path for a given destination. NTU IM OPLab

44. LQSR(Link Quality Source Routing) • It combines link-state proactive routing with the reactive strategy from ad hoc networks. • It uses a complete view of the network topology to compute shortest paths. • Nevertheless, it uses a route discovery procedure as in reactive protocols to reduce routing overhead, which may become high because of medium instability and user mobility. NTU IM OPLab

45. SrcRR • It uses only a discovery procedure similar to reactive protocols to update the routing information of the traversed links, reducing control overhead. • It computes routes using a reduced view of the network. NTU IM OPLab

46. SrcRR • Every node running SrcRR maintains a link cache, which tracks the ETX metric values for links it has heard about recently. • Whenever a change is made to the link cache, the node locally runs Dijkstra's weighted shortest-path algorithm on this database to find the current, minimum-metric routes to all other nodes in the network. • To ensure only fresh information is used for routing, if a link metric has not been updated within 30 seconds it is dropped from the link cache. NTU IM OPLab

47. SrcRR • When a node wants to send data to a node to which it does not have a route, it floods a route request. • When a node receives a route request, it appends its own node ID and current ETX metric from the node from which it received the request, and rebroadcasts it. • A node will always forward a given route request the first time it receives it. • If it receives the same route request again over a different route, it will forward it again if the accumulated route metric is better than the best metric it has forwarded so far. • This ensures that the target of the route request will receive the best routes. NTU IM OPLab

48. SrcRR • When a node receives a route request for which it is the target, it reverses the accumulated route and uses this as the source-route for a route reply. • When the original source node receives this reply, it adds each of the links to its link cache, and then source-routes data over the minimum-metric path to the destination. NTU IM OPLab

49. SrcRR • When a SrcRR node forwards a source-routed data packet, it updates its entry in the route to contain the latest ETX metric for the link on which it received the packet. • All query and data packets contain ETX metrics for the links they have traversed so far. • Any node that receives such a packet (including forwarding nodes) copies those metrics to its link cache. • As with all changes to the link cache, this prompts recomputation of all the best routes using Dijkstra's algorithm. • This allows the source and destination to maintain up-to-date link caches. NTU IM OPLab

50. SrcRR • Each data packet includes a field to hold one extra link metric, and this allows the source and destination to learn of the existence and metric of some alternate links. • Both LQSR and SrcRR implement route discovery procedures using source routing and ETX. NTU IM OPLab