We consider a multiple access relay network where multiple sources send independent data to a single destination through multiple relays, which may inject falsified data into the network. To detect the malicious relays and discard (erase) data from them, tracing bits are embedded in the information data at each source node. In addition, parity bits are added to correct the errors caused by fading and noise. When the total amount of redundancy, tracing bits plus parity bits, is fixed, an increase in parity bits to increase the reliability requires a decrease in tracing bits, which leads to a less accurate detection of malicious behavior of relays, and vice versa. We investigate the tradeoff between the tracing bits and the parity bits in minimizing the probability of decoding error and maximizing the throughput in multisource, multirelay networks under falsified data injection attacks. The energy and throughput gains provided by the optimal allocation of redundancy and the tradeoff between reliability and security are analyzed.

SIFT has been proven to be the most robust local invariant feature descriptor. SIFT is designed mainly for gray images. However, color provides valuable information in object description and matching tasks. Many objects can be misclassified if their color contents are ignored. This paper addresses this problem and proposes a novel colored local invariant feature descriptor. Instead of using the gray space to represent the input image, the proposed approach builds the SIFT descriptors in a color invariant space. The built Colored SIFT (CSIFT) is more robust than the conventional SIFT with respect to color and photometrical variations. The evaluation results support the potential of the proposed approach.

Cooperative relaying is gaining a significant attention in that, intermediate relay nodes assist the source nodes to enhance the overall network efficiency. In network coded cooperative communications, the relay node linearly combines the data received from the sources and forward the linear combination to the destination. In this paper, we present the maximum likelihood decoding scheme for a network composed of two sources, single relay, and single destination. We derive a closed form expression for upper bound on the word error probability. The simulation results show that the analytical upper bound is very tight especially at higher values of the SNR. The results also show that there exists an error floor in the error probability. Therefore, the closed form expression of the upper bound is used to study the reasons of this error floor and how to mitigate it.

We analyze the probability of decoding error in a multi-source, multi-relay wireless network, in which the adversary may inject falsified data through captured relay nodes. To detect malicious relay nodes and discard (erase) the data from those relay nodes, tracing bits are embedded in the information data at each source node. Parity bits are also added to the information data to correct the errors caused by the the channel impairments such as fading and noise. We analyze the tradeoff between the tracing bits and the parity bits in minimizing the probability of decoding error in multi-source, multi-relay networks under falsified data injection attacks.

NULL

NULL

NULL