Poster Abstract: Alternative Path Creation vs Data Rate Reduction for Congestion Mitigation in Wireless Sensor Networks Charalambos Sergiou and Vasos Vassiliou
Department of Computer Science, University of Cyprus
sergiou,
[email protected]
ABSTRACT Network Congestion is a highly undesirable situation for every type of network. Especially in low powered Wireless Sensor Networks (WSNs), congestion can be proven critical for the network’s proper operation. In case of congestion in a WSN, the network is programmed to react, either by reducing the data rate of the sources or by creating multiple routing paths to the sink, thus avoiding the network’s congested point. In this paper we perform a comparison of these two techniques, comparing one algorithm of each category. Results depict the advantages and disadvantages of each category.
Table 1: Simulation Parameters X distance (m) 1000 Path Loss Coeff. Y distance (m)
1000
3.5
Node CPU(MHz)
4
TX Power (mW)
60
Data packet
128 bytes
RX Power (mW)
60
Control Packet
50 bytes
Idle Power (mW)
5
MAC Prot.
CSMA
Sensitivity (dBm)
-81
Initial Energy
1 Joule
compare two algorithms (one of each congestion mitigation category) trying to extract conclusions for the advantages Categories and Subject Descriptors and disadvantages of each technique. Specifically we use the C.2.1 [Computer Communication Networks]: Network HTAP (Hierarchical Tree Alternative Path) [2] algorithm Architecture and Design—Wireless Networks; C.2.2 [Computer that creates alternative paths in case of congestion in order Communication Networks]: Network Protocols—Routto avoid the congested paths and Sen-TCP [1] algorithm, ing protocol an algorithm that reacts to congestion through the reduction of the sources’ data rate. In a related work [3] we have General Terms also examined how congestion is affected by the actual node placement in the sensing area. Algorithms, Performance, Design, Reliability
Keywords Topology Control, Congestion Control, Energy Utilization
1.
INTRODUCTION
A special category of WSNs are the event-based networks. In event-based networks data packets are produced only upon the observation of a particular event that satisfies a pre-specified condition. Due to the nature of these environments, sudden traffic increase occurs when the monitoring event is happening. This high generation of data packets is usually uncontrolled and often leads to congestion. Congestion drives the network into an unstable state. In this ˇ buffers overflow [4] resulting in random drops state nodesS of data packets. Dropped packets are a major handicap for these networks since they result in severe energy consumption. In the case that no countermeasures are taken the power of congested nodes can be exhausted leading to the creation of ”dead paths” (holes) in the network. There are several algorithms in literature designed to counter, control or avoid congestion. The majority of these algorithms use two techniques to mitigate congestion: by source data rate reduction or by multiple path creation. In this paper we Copyright is held by the author/owner(s). IPSN’10, April 12–16, 2010, Stockholm, Sweden. ACM 978-1-60558-955-8/10/04.
2.
DESCRIPTION OF ALGORITHMS
The HTAP algorithm [2] is a distributed and scalable algorithm consisting of four major parts. • Flooding with level discovery functionality • The Hierarchical Tree Algorithm • Alternative Path Creation Algorithm • Handling of Powerless (Dead Nodes) SenTCP [1] is an open-loop hop-by-hop congestion control protocol with two special features: • It jointly uses average local packet service time and average local packet inter-arrival time in order to estimate current local congestion degree in each intermediate sensor node. • It uses hop-by-hop congestion control
3. PERFORMANCE EVALUATION 3.1
Simulation Environment
To perform comparison between the algorithms concerning their ”congestion” performance, simulations have been conducted according to the parameters of Table 1.
25 HTAP Sen TCP
0.035
0.03
0.025
0.02
15
10
5
0.015
0.01 100
HTAP Sen− TCP
20 Average Packet Drops
Average Node Energy Consumption (J)
0.04
150
200
250 300 350 Number of Nodes
400
450
0
500
1
2
3
4
5 6 7 Simulation Time (s)
8
9
10
Figure 3: Packet Drops
Figure 1: Average Node Energy Consumption 800 500 Average Data Rate (Packets/Second)
Average Number of Power Exhausted Nodes
700 HTAP Sen TCP
450 400 350 300 250 200 150
500 400 300 200 100
100 0 50 100
150
200
250 300 350 Number of Nodes
400
450
Simulation Results
The first parameter we compare is the Average Node Energy Consumption. As it is presented in Fig. 1, HTAP consumes more energy per node compared to Sen-TCP. The reason is the fact that it employs a big percentage of nodes in order to forward the data, since the packets are routed through alternative paths in case of congestion. On the other hand, when Sen-TCP is employed, the average node energy consumption is less, since there are many nodes which remain in full power, due to the fact that Sen-TCP always uses the shortest path to forward data. Fig. 2 presents results on the network lifetime. In this case simulations are run until the network becomes unable to forward packets from the source to the sink. It is clear that HTAP, through the creation of alternative paths, is able to balance the traffic among nodes and to extend the network lifetime. Sen-TCP on the other hand will revert to alternate paths only after the shortest available path is power exhausted. This gives it a disadvantage over HTAP. Fig. 3 depicts the number of packets drops during 10 seconds of simulation time. This graphs proves that both algorithms maintain low packet drops. Initially Sen-TCP looses bigger number of packets than HTAP but during the evolution of time, improves its performance due to the fact that it controls the data rate. Thus, it maintains its losses to nearly null levels. Finally, Fig. 4 presents the Data rate on which sources are injecting packets in the network. HTAP maintains a stable data rate while Sen-TCP evolves its data rate based on packet losses.
4.
1
2
3
4
500
CONCLUSION In this paper we compare the ”congestion” performance
5 6 7 Simulation Time (s)
8
9
10
Figure 4: Average Data Rate
Figure 2: Number of Power Exhausted Nodes
3.2
HTAP Sen TCP
600
of two algorithms that counter congestion in different ways. HTAP uses alternative paths in order to ”by pass” the congested area and Sen-TCP reduces the source’s data rate to mitigate congestion. Simulation results present the advantages and disadvantages of these techniques. Alternative path creation algorithms present bigger network lifetime while keeping the data rate stable, whereas data rate reduction algorithms present less power consumption per node and minimal packet drops.
5.
ACKNOWLEDGMENTS
This work has been conducted under the European Union Project GINSENG funded under the FP7 Program (FP7/20072013) grant agreement no 224282.
6.
REFERENCES
[1] B. L. C. Wang, K. Sohraby. Sentcp: a hop-by-hop congestion control protocol for wireless sensor networks. IEEE INFOCOM (Poster Paper), March 2005. [2] C. Sergiou, V. Vassiliou, and A. Pitsillides. Reliable data transmission in event-based sensor networks during overload situation. In WICON ’07: Proceedings of the 3rd international conference on Wireless internet, Austin, Texas, October 2007. [3] V. Vassiliou and C. Sergiou. Performance study of node placement for congestion control in wireless sensor networks. In 3rd International Conference on New Technologies, Mobility and Security (NTMS), Cairo, Egypt, Dec., 2009. [4] A. Woo and D. E. Culler. A transmission control scheme for media access in sensor networks. In Proceedings of the 7th annual international conference on Mobile computing and networking (MobiCom’01), pages 221–235, New York, NY, USA, 2001. ACM Press.
