Overview:

In this project, we investigate the benefits of using the peer-to-peer network model in cellular wireless packet data networks. First, we study the performance trade-offs between conventional cellular networks (cellular network model) and multi-hop ad-hoc networks (peer-to-peer network model). We find that while the peer-to-peer network model performs better in terms of throughput and power consumption, they suffer from unfairness and severe performance degradation due to mobility and traffic locality. The trade-offs preclude the adoption of either of the network models as a clear solution for future wireless communication systems.

Therefore, we explore a new hybrid wireless network model called Sphinx that has a dual mode of operation. Sphinx uses a peer-to-peer network model in tandem with the conventional cellular network model to achieve higher throughput and lower power consumption. At the same time, Sphinx avoids the typical pitfalls of a pure peer-to-peer network model including unfair resource allocation and throughput degradation due to mobility and traffic locality. We show through simulation results that Sphinx outperforms the conventional cellular network model in terms of throughput and power consumption, and achieves better fairness and resilience to mobility and traffic locality than the peer-to-peer network model.

Results / Status:

Figure 1	Figure 2

In Figures 1 and 2, we show the performance of the hybrid network model in terms of the impact on mobility and traffic locality. The simulation results are obtained using the ns-2 network simulator. We use a topology with 100 mobile stations randomly distributed in a 1500m by 1500m cell. We use 50 TCP flows and measure the average throughput as the performance metric.

Figure 1 shows the impact of mobility on the throughput. All flows have the source and destination randomly chosen from the 100 mobile stations. Mobile stations randomly move within the cell using the way-point mobility model. We observe that the throughput of the peer-to-peer model suffers for higher mobility rates while the throughput of the cellular model remains unaffected because the mobility is only intra-cell. The performance of the hybrid model tracks the performance of the peer-to-peer model during lower speeds. However, for the higher speeds, more flows experience throughput degradation and are switched to the cellular mode. Hence, the performance of the hybrid model tracks the performance of the cellular model for higher speeds and hence does not show the throughput degradation exhibited by the peer-to-peer model.

Figure 2 shows the impact of traffic locality on the throughput. A local flow is defined as one that has both the source and destination inside the same cell, while a non-local flow is one that has either the source or the destination outside the cell. We vary the percentage of non-local flows in the network and compare the performance of the three network models. As in the case of mobility, the performance of the hybrid model tracks that of the peer-to-peer model for lower percentages when the peer-to-peer model is significantly better than the cellular model, and switches over to the cellular model for the extreme cases where most of the flows are non-local and hence the cellular model performs better than the peer-to-peer model.

Publications & Presentations:

• H.-Y. Hsieh and R. Sivakumar,
  ``Performance Comparison of Cellular and Multi-hop Wireless Networks: A Quantitative Study.''
  ACM International Conference on Measurement and Modeling of Computer Systems (SIGMETRICS), Cambridge, MA USA, June 2001.
• H.-Y. Hsieh and R. Sivakumar,
  ``Towards a Hybrid Network Model for Wireless Packet Data Networks.''
  IEEE Symposium on Computers and Communications (ISCC), Taormina, Italy, July 2002.
• H.-Y. Hsieh and R. Sivakumar,
  ``A Hybrid Network Model for Cellular Wireless Packet Data Networks.''
  IEEE Global Communications Conference (GLOBECOM), Taipei, Taiwan, November 2002.
• H.-Y. Hsieh and R. Sivakumar,
  ``Internetworking WWANs and WLANs in Next Generation Wireless Data Networks.''
  International Conference on 3G Wireless and Beyond, San Francisco, CA USA, May 2002.
• H.-Y. Hsieh and R. Sivakumar,
  ``On Using the Ad-hoc Network Model in Cellular Packet Data Networks.''
  ACM International Symposium on Mobile Ad Hoc Networking and Computing (MOBIHOC), Lausanne, Switzerland, June 2002.
• H.-Y. Hsieh and R. Sivakumar,
  ``On Using Peer-to-Peer Communication in Cellular Wireless Data Networks.'' Accepted for publication in IEEE Transactions on Mobile Computing, 2004.

Software Downloads:

People:

• Hung-Yun Hsieh (Alumnus)
• Raghupathy Sivakumar (Professor)

References & Related Work:

• A-GSM
  G. Aggelou and R. Tafazolli, "On the Relaying Capacity of Next-Generation GSM Cellular Networks," IEEE Personal Communications Magazine, vol. 8, no. 1, pp. 40-47, February 2001.
• ODMA
  3GPP TSG-RAN, "Opportunity Driven Multiple Access," 3GPP Technical Report, 3G TR 25.924, v1.0.0, December 1999.
• IR
  T. Harrold and A. Nix, "Intelligent Relaying for Future Personal Communication Systems," IEE Colloquium on Capacity and Range Enhancement Techniques for the Third Generation Mobile Communications and Beyond, February 2000.
• iCAR
  C. Qiao and H. Wu, "iCAR: An Intelligent Cellular and Ad-hoc Relay System," Proceedings of IEEE IC3N, Las Vegas, NV USA, October 2000.
• MCN
  Y.-D. Lin and Y.-C. Hsu, "Multihop Cellular: A New Architecture for Wireless Communications," Proceedings of IEEE INFOCOM, Tel-Aviv, Israel, March 2000.
• MADF
  X. Wu, S.-H. Chan and B. Mukherjee, "MADF: A Novel Approach to Add an Ad-hoc Overlay on a Fixed Cellular Infrastructure," Proceedings of IEEE WCNC, Chicago, IL USA, September 2000.