Symbol-level information extraction : from bits to symbols in wireless network cross-layer design

Abstract

At ever-increasing rates, wireless networks are becoming an indispensable part of people's daily life due to the low cost in the deployment of networking infrastructures and the high availability to access the Internet. Current wireless technologies use bits to deliver both the digital contents and packets' information, e.g. MAC address or packet transmission duration, which require the packets to be fully received and correctly decoded. However, for most receivers in the networks, they only need packets' information to adjust their behaviors even if they suffer the bad channel conditions. In this thesis, I mainly focus on the cross-layer design of symbol-level information extraction which enables the vital information to be delivered between the transmitters and receivers at symbol-level. I further take advantage of that symbol-level information to design the cross-layer structure to deal with three popular research topics in wireless communications: the hidden terminal problem, the energy inefficiency of the packet overhearing problem and communications security (COMSEC) problem. Hidden terminals are typical interference sources that can significantly reduce the throughput of a wireless network if it adopts the CSMA/CA (Carrier Sense Multiple Access with Collision Avoidance) MAC protocol. Normally, the standard RTS/CTS (Request to Send/Clear to Send) mechanism is deployed to solve this hidden terminal problem. However, the standard RTS/CTS would fail to silence all the hidden terminal, as in the real world, the CTS packets might not be correctly received all the time due to either the CTS packets are unable to be decoded at remote hidden terminals or the CTS packets are collided with other packets at the hidden terminals. To solve the two drawbacks of the CTS packets, the RTS/S-CTS mechanism is proposed, which uses the symbol-level information extraction mechanism to deliver the NAV time information. The S-CTS frame can be correctly detected from collisions and by remote hidden terminals. Thus, the NAV time information which is contained in the S-CTS frame can be used to silence those hidden terminals. A testbed of RTS/S-CTS with GNURadio/USRP2 software radio is built to demonstrate its feasibility and extensive ns-2 simulations are conducted to evaluate its performance. The simulations results show that the RTS/S-CTS can significantly improve the throughput in the random topology network scenario compared with the standard RTS/CTS. In the energy inefficiency of the packet overhearing problem, I first reveal that the energy waste on the packet overhearing accounts for the majority of the energy inefficiency of wireless devices in high traffic wireless LANs by analyzing the real world traffic traces. Though there are many existing approaches trying to reduce this energy inefficiency, none of them can avoid the energy waste on the packet overhearing in high traffic wireless LANs effectively. Thus, I propose a novel SASD (Sample-Address Sample-Duration) scheme that uses the symbol-level information extraction mechanism to deliver the MAC address and packet transmission duration information at the PHY layer. The SASD enables the wireless devices to discern the information under the energy-saving downclocking mode through the SASD Detection and Identification decoder. Consequently, the devices which are not the intended receiver of the packet can switch to the sleeping mode to save the energy cost on the packet overhearing. The extensive hardware experiments and simulation results show that the SASD can greatly outperform the existing approaches in the high traffic wireless LAN scenario. In the wireless communications security problem, I explore the feasibility of symbol obfuscation to defend against the passive eavesdropping attack and fake packet injection attack during the wireless communications. I propose a novel Multiple Inter-symbol Obfuscation (MIO) scheme, which utilizes a set of artificial noisy symbols, which called "symbols key", to obfuscate the original data symbols in the PHY layer. As the symbols key information can only be extracted and verified by the legitimate receiver, the eaves-dropper can hardly decrypt the obfuscated symbols from the eavesdropped packets, and the fake packet can be easily checked out. Thus, the MIO can effectively enhance the wireless communications security. Compared with other communications security methods, the mathematical analysis proves that the MIO can provide the better performance on computational secrecy against the fake packet injection attack. Moreover, MIO can provide an easier way to achieve information-theoretic secrecy against the passive eaves-dropping attack. To sum up, by investigating the symbol-level information extraction mechanism to solve the problems, we show the feasibility of delivering vital information at the symbol level. Moreover, this symbol-level information extraction mechanism opens up a new dimension to convey the information from the bit level to the symbol level.