Inspired by the challenges of designing a robust, and undetectable covert channel, in this paper we introduce a design methodology for timing covert channels that achieve provable polynomial-time undetectability. This means that the covert channel can not be detected by any polynomial-time statistical test that analyzes the samples of the covert traffic and the legitimate traffic. The proposed framework is based on modeling the covert channel as a differential communication channel, and the formulation for modulation/demodulation processes that are derived according to the communication model. The proposed scheme incorporates a trellis structure in modulating the covert message. The trellis structure is also used at the covert receiver to perform iterative demodulation/decoding of the covert message that significantly enhances the channel reliability. In addition, the paper presents an adaptive modulation strategy that improves the channel robustness without compromising the stealthiness of the channel. The combination of the adaptive modulation and the trellis structure gives the covert channel considerable flexibility and low error rate at the covert receiver. In fact, performance analysis of the channel reveals that the proposed covert communication scheme withstands extremely high levels of network noise and adversarial disruption, while it maintains an outstanding undetectability level and covert rate.

In this paper, the use of structural behavior of communication protocols in designing new covert channels is investigated. In this way, a new covert transmitter is designed based on a modified CSMA protocol that enables the transmitter to embed a covert message in its overt traffic. The proposed scheme provides high covert rate without compromising the stealthiness of the channel.

Dynamic source routing (DSR, introduced in 1996) is one of the most frequently used routing protocols for mobile ad-hoc networks (MANETs). Numerous MANET protocols were created based on DSR's algorithm. In addition to inheriting the overall performance specifications of DSR, these MANET protocols are designed to perform optimised for specific functionality. To name a few, these functionalities include: hierarchical routing, security-aware routing and multipath routing. Such flavoured DSR schemes (X-DSR) are often compared against the original DSR protocol through simulation results. The purpose of this survey is to first introduce DSR in detail, discuss most of the DSR flavours, point out their specific features, and to present a complete survey of the analyses given in the current literature against the original DSR protocol. Following this in-depth discussion, we introduce an X-DSR-aware management architecture, which utilises a multilayer scheme that imports parameters from different layers (network, data-link and physical) and performs current network condition matching compared to the closest pre-defined network condition groups. The output of such a match is the selection of the most optimal routing protocol, which satisfies most of the criteria of the pre-defined condition group.

In many applications of sensor networks, the sink needs to keep track of the history of sensed data of a monitored region for scientific analysis or supporting historical queries. We call these historical data a time series of value distributions or snapshots. Obviously, to build the time series snapshots by requiring all of the sensors to transmit their data to the sink periodically is not energy efficient. In this paper, we introduce the idea of gradient boundary and propose the gradient boundary detection (GBD) algorithm to construct these time series snapshots of a monitored region. In GBD, a monitored region is partitioned into a set of subregions and all sensed data in one subregion are within a predefined value range, namely, the gradient interval. Sensors located on the boundaries of the subregions are required to transmit the data to the sink and, then, the sink recovers all subregions to construct snapshots of the monitored area. In this process, only the boundary sensors transmit their data and, therefore, energy consumption is greatly reduced. The simulation results show that GBD is able to build snapshots with a comparable accuracy and has up to 40 percent energy savings compared with the existing approaches for large gradient intervals.

In this paper, we propose a novel energy-efficient approach, a localized routing tree (LRT) coupled with a route redirection (RR) strategy, to support various types of queries. LRTs take care of the sensors near the sink and reduce the energy consumption of these sensors, and RR reduces the energy cost of data receptions. Compared to the existing approaches, simulation studies show that LRT together with RR has significant improvement on the query capacity.

Hierarchical routing is an effective way to solve the scalability problem in flat networks. The optimal total routing table (RT) size is affected by four parameters: the total number of nodes in a network hierarchy levels, the number of clusters at each level, and cluster size distribution. An optimal solution of total RT size was given in L. Kleinrock et al. (1977). However, the optimal solution was based on a major assumption: all nodes in the network have equal size RTs. It is not clear whether the optimal results stated in L. Kleinrock et al. (1977) still hold without this assumption. We provide the general integer solution of optimal RT sizes without this assumption. In addition, we will show how the total number of nodes, the number of hierarchical levels, and the number of the highest-level clusters affect the total RTsize. Moreover, some important properties of the two-level cluster structure are extensively addressed, namely the impact of the variance of cluster size distribution on intra-cluster update costs and the RTsize.

Hierarchical routing is used to reduce routing update costs, and clustering algorithms partition networks into multilevel cluster structures. In hierarchical networks, routing performance is greatly affected by the sizes and structures of partitioned clusters. Two existing clustering algorithms are based on the concepts of the lowest ID and the maximum degree of nodes. In this paper, we present a new clustering algorithm based on the concept of the variable degree of nodes. Our new algorithm produces clusters with the low cluster variance in size, and the number of the clusters is close to the optimal value that leads to the smallest total routing table size. The performance of the algorithm is evaluated by using simulated ad hoc networks.

The authors describe a VLSI Galois field processor and how it can be applied to the implementation of elliptic curve groups. They demonstrate the feasibility of constructing very fast, and very secure, public key systems with a relatively simple device, and the possibility of putting such a system on a smart card. The registers necessary to implement the elliptic curve system will require less than 1 mm/sup 2/ (or less than 4%) of the area available on the card.>

In this paper we examine the development of a high-speed implementation of a system to perform exponentiation in fields of the form GF(2 n ). For sufficiently large n, this device has applications in public-key cryptography. The selection of representation and observations on the structure of multiplication have led to the development of an architecture which is of low complexity and high speed. A VLSI implementation has being fabricated with measured throughput for exponentiation for cryptographic purposes of approximately 300 kilobits per second.

The problem of ensuring recoverability of encrypting data in a file storage system is examined. In this situation, the original data (plaintext) are removed from the system after encryption. In the event of errors either in the initial processing or in the storage of the data, the file may be rendered unrecoverable. A model of a file storage system that incorporates errors in these areas is developed. A two-stage coding system involving error-correcting codes and interleaving is introduced. The performance of this system is analyzed for various sizes of encryption blocks and error-correcting codes.>

Network performance is investigated at an isolated node which we call the switching node (SN). An equal (outgoing) link utilization constraint is introduced to enhance the robustness of a minimum delay routing algorithm applied to the SN in the network. Input traffic to the SN is grouped into classes in accordance with the radial distance from the SN to their destinations. It is shown that under an equal link utilization constraint a four-class traffic admission scheme displays a smaller probability of buffer overflow compared to an unconstrained routing algorithm. Also, with the equal link utilization constraint, the buffer overflow characteristics are similar for both balanced and unbalanced loads. It is shown that the four-class traffic admission model exhibits an "almost" deadlock free characteristic, e.g., with a total buffer size of 4, the probability of deadlock is about 10-10.