Overview:

The rapid growth in the use of smartphones and WiFi enabled devices has exposed the limitations of traditional wireless networking solutions. This project focuses on three key performance metrics: link quality, security and spatial reuse in high-density wireless LAN deployments. To address the challenges in each of these problems, we leverage smart antennas and cooperation across access points to achieve a variety of network-level benefits.

Impact:

Wireless LAN deployments now exceed a million all over the world, with global revenues running into several tens of billions of dollars. However, practical performance in WLANs is especially poor, with unreliable, low throughput and insecure links.

The key contributions of this work are novel approaches to enhance security and throughput performance using smart antennas and cooperative transmission. We introduce the notion of physical-space security, where the knowledge of existence of information and the content are limited from eavesdroppers at the spatial level without relying on computational hardness. We also introduce the notion of topology aware cooperative coding and practical beamforming algorithms to realize successful information transfer even with traditionally interfering links.

Concept:

A smart WiFi architecture, illustrated in Figure 1, consists of beamforming and cooperatively coded links that (i) minimize the exposure region to eavesdroppers (Aegis [1]), (ii) increase the Signal To Noise Ratio (SNR) at receivers by forming robust and minimally interfering beams (Symbiotic Coding [2] and LSR [3]) and (iii) include adaptation algorithms to adapt beams, coding and modulation in response to varying network conditions. The key strategies used are:

1. Information deprivation - uses flexible beamforming approaches to prevent information signals from reaching the eavesdropper and spatial diversity to ensure that an eavesdropper at any location cannot get the complete information.
2. Information concentration – leverage the high density of access points to cause independent information components to collide and reduce the ability of the eavesdroppers to decode.
3. Information coordination – leverage topological and spatial diversity across access points to code information such that harmful (undecodable) symbol combinations are avoided at the receivers.

Figure 1: Architecture

Results:

In simulations and in real-life experiments, Aegis provides robust security over competitive approaches as illustrated in Figure 2. Similarly coding and LSR improve performance in high-density deployments as illustrated in Figure 3 and Figure 4.

Figure 2: Aegis performance

Figure 3: Symbiotic Coding

Figure 4: Scalability of LSR

Publications:

1. S. Lakshmanan, C.-L. Tsao, and R. Sivakumar, "On Coding Concurrent Transmissions in Wireless Networks," Poster Presentation, ACM International Conference on Mobile Computing and Networking (MOBICOM), Beijing, China, Spet. 20-25, 2009.
2. S. Lakshmanan, K. Sundaresan, S. Rangarajan, and R. Sivakumar, "Practical Beamforming based on RSSI Measurements using Off-the-shelf Wireless Clients," ACM SIGCOMM conference on Internet measurement (IMC), Chicago, Illinois, November 4-6, 2009.
3. S.Lakshmanan, C.-L. Tsao, and R. Sivakumar, "Aegis: Physical Space Security for Wireless Networks With Smart Antennas," IEEE/ACM Transactions on Networking, vol. 18, no. 4, pp. 1105-1118, August 2010.
4. S. Lakshmanan, C.-L. Tsao, R. Sivakumar, and K. Sundaresan, "Securing Wireless Data Networks against Eavesdropping using Smart Antennas," International Conference on Distributed Computing Systems (ICDCS), Beijing, China, June 2008.

People:

• Sriram Lakshmanan (Student)
• Sandeep Kakumanu (Student)
• Shruti Sanadhya (Student)
• Raghupathy Sivakumar (Professor)