Osameh M. Al-Kofahi

Network coding-based protection of many-to-one wireless flows

This paper addresses the problem of survivability of many-to-one flows in wireless networks, such as wireless mesh networks (WMNs) and wireless sensor networks (WSNs). Traditional protection schemes are either resource-hungry like the (1+1) protection scheme, or introduce a delay and interrupt the network operation like the (1 : N) protection scheme. In this paper, we present a network coding-based protection technique that overcomes the deficiencies of the traditional schemes. We derive and prove the necessary and sufficient conditions for our solution on a restricted network topology. Then we relax these connectivity requirements and show how to generalize the sufficient and necessary conditions to work with any other topology. We also show how to perform deterministic coding with {0,1} coefficients to achieve linear independence. Moreover, we discuss some of the practical considerations related to our approach. Specifically, we show how to adapt our solution when the network has a limited min-cut; we therefore define a more general problem that takes this constraint into account, which prove to be NP-complete. Furthermore, we discuss the decoding process at the sink, and show how to make use of our solution in the upstream communication (from sink to sources). We also study the effect of the proposed scheme on network performance. Finally, we consider the implementation of our approach when all network nodes have single transceivers, and we solve the problem through a greedy algorithm that constructs a feasible schedule for the transmissions from the sources.

Efficient and Agile 1+N Protection

This paper introduces an efficient implementation of the network coding-based 1+N protection. The strategy provides proactive protection to N link-disjoint full-duplex connections against single link failures. The implementation is efficient and uses a tree shaped minimum cost protection circuit. The protection circuit carries linear combinations of data units originally transmitted on the working circuits, and these linear combinations can be used to recover data units lost due to failures. This recovery is carried out with the assistance of one node on the protection tree, which is chosen to reduce the recovery time. This protection technique requires the same amount of protection resources used by 1:N protection, where the protection circuit is link disjoint from the protected connections. The paper also makes other contributions. It introduces an Integer Linear Program (ILP) formulation to evaluate the cost of protection using this technique, and compares it to the cost of 1+1 protection. The comparison shows that a significant saving in cost can be achieved, while recovering from failures within a short time. The performance of this scheme is further evaluated using an OPNET-based simulation, where it was shown that the recovery time conforms to acceptable industry standards. Availability analysis is also conducted.

Survivability strategies in multihop wireless networks [Accepted From Open Call]

Survivability is an important network characteristic that provides a certain level of data delivery guarantees. The degree of survivability is usually governed by the data transfer mechanism or protocol that delivers data from source to destination. In this article we survey and discuss a variety of survivability issues, challenges, and mechanisms in multihop wireless networks. Unlike some previous surveys, we do not focus only on multipath routing techniques. We try to cover a broader spectrum of survivability techniques in the literature. Moreover, we discuss new directions in survivability that use the network coding technique in order to achieve a better degree of scalability, which is usually an issue in most survivability techniques, especially in wireless networks.

Network protection codes: Providing self-healing in autonomic networks using network coding

Agile recovery from link failures in autonomic communication networks is essential to increase robustness, accessibility, and reliability of data transmission. However, this must be done with the least amount of protection resources, while using simple management plane functionalities. Recently, network coding has been proposed as a solution to provide agile and cost efficient self-healing against link failures, in a manner that does not require data rerouting, packet retransmission, or failure localization, hence leading to simple control and management planes. To achieve this, separate paths have to be provisioned to carry encoded packets, hence requiring either the addition of extra links, or reserving some of the resources for this purpose. In this paper we introduce self-healing strategies for autonomic networks in order to protect against link failures. The strategies are based on network coding and reduced capacity, which is a technique that we call network protection codes (NPC). In these strategies, an autonomic network is able to provide self-healing from various network failures affecting network operation. Also, network protection codes are extended to provide self-healing from multiple link failures in autonomic networks. Although this leads to reducing the network capacity, the network capacity reduction is asymptotically small in most cases of practical interest. We provide implementation aspects of the proposed strategies, derive bounds and show how to construct network protection code. The paper also develops an Integer Linear Program formulation to evaluate the cost of provisioning connections using the proposed strategies, and uses results from this formulation to show that it is more resource efficient than 1+1 protection. A simulation study to evaluate the recovery times, and the buffering requirements due to network coding is also conducted using the OPNET simulator.

Scalable redundancy for sensors-to-sink communication

In this paper, we present a new technique that uses deterministic binary network coding in a distributed manner to enhance the resiliency of sensor-to-base information flow against packet loss. First, we show how to use network coding to tolerate a single packet loss by combining the data units from sensor nodes to produce k+1 combinations such that any k of them are solvable. After that, we extend the solution to tolerate multiple losses. Moreover, we study the coding efficiency issue and introduce the idea of relative indexing to reduce the coding coefficients overhead. To tolerate node or link failures, we introduce a simple routing protocol that can find maximally disjoint paths from the k sensor nodes to the base station (BS). We study the relationship between the probability of successful recovery of all data units at the BS, and the number of sources protected together taking into consideration their hop distance from the BS. From this study, we can decide on the appropriate number of sources to be protected together, so that the probability of successful recovery is higher than a certain threshold. Finally, we show through a simulation study that our approach is highly scalable and performs better as the network size increases.

Network Coding-Based Protection of Many-to-One Flow Networks

Survivability of many-to-one flow networks such as wireless mesh networks (WMNs) and wireless sensor networks (WSNs) is an important issue that has not received enough attention in the literature. Traditional proactive and reactive protection schemes are either resource-hungry like the (1 + 1) protection scheme, or introduce a delay and interrupt the network operation like the (1 : N) protection scheme. In this paper, we present a novel approach that relies on network coding to provide protection to many-to-one flows as in WMNs or WSNs at the speed of proactive protection, but at the cost of reactive protection. We derive and prove the necessary and sufficient conditions for our solution on a restricted network topology. Then we introduce three generalizations of our problem, and provide a heuristic and a mixed integer linear program (MILP) to solve one of them. We also show how to perform deterministic coding with {0,1} coefficients to achieve linear independence. Finally, we discuss some practical considerations related to our approach, and define a more general problem that takes these considerations into account. We also formulate this general problem as an MILP.

Scalable Redundancy for Sensors-to-Sink Communication

In this paper, we present a new technique that uses deterministic binary network coding in a distributed manner to enhance the resiliency of sensor-to-base information flow against packet loss. First, we show how to use network coding to tolerate a single packet loss, and then we extend the solution to tolerate multiple losses. Moreover, we study the coding efficiency issue and introduce the idea of relative indexing to reduce the coding coefficients overhead. Finally, we show through a simulation study that our approach is highly scalable and performs better as the network size and/or number of sources increases.

Taking advantage of jamming in wireless networks

There has been an increasing interest in jamming-based solutions to improve the performance of wireless networks. In such schemes, users would deliberately jam the channel, with a special signal, to improve the efficiency of different operations related to wireless networks especially when high performance is required like in contention resolution, QoS support, and statistics estimation. Thus, in this paper we seek to describe jamming and identify positive employments of the jam signals in such networks. We identify different areas where jamming signals are exploited to improve the performance and to solve problems that may arise in wireless networks. We also attempt to categorize different protocols in some of these areas. We provide a comprehensive survey of different proposals that exploit jamming to efficiently perform operations and solve problems in a wireless network. We highlight challenges, requirements, advantages, and disadvantages of jamming.

Toward an optimal 1+N protection strategy

This paper introduces an implementation of the network coding-based generalized 1+N protection technique presented earlier by the author in [1] to protect against single link failures. Instead of using two protection circuits for a group of connections which are to be protected together as in [1], only one protection circuit is used, which takes the form of a tree. The protection circuit carries linear combinations of the data units originally transmitted on the working circuits, and these linear combinations can be used to recover lost data due to link failures. This recovery is carried out with the assistance of one node on the protection tree, which is chosen to reduce the recovery time. Moreover, unlike the scheme in [1] which protects unidirectional connections, this scheme is used to protect bidirectional connections. This protection technique requires exactly the same amount of protection resources used by 1:N protection, and can therefore be considered as a step towards achieving optimal 1+N protection. The paper also makes a number of other contributions. It introduces an integer linear program (ILP) formulation to evaluate the cost of protection using this technique, and compares it to the cost of 1+1 protection. The comparison shows that a significant saving in cost can be achieved, while still recovering from failures within a short time.

Max-Flow Protection Using Network Coding

In any communication network, the maximum number of link-disjoint paths between any pair of communicating nodes, S and T, is limited by the S-T minimum link-cut. Multipath routing protocols have been proposed in the literature to make use of these S-T paths in enhancing the survivability of the S-T information flow. This is usually done by using a subset of these paths to forward redundant data from S to T. Therefore, this enhancement in survivability reduces the useful S-T information rate. In this paper we present a new way to enhance the survivability of the S-T information flow without compromising the maximum achievable S-T information rate, which is especially important in wireless networks where wireless resources are limited. To do this, bottleneck links (in the min-cut) should only forward useful information, and not redundant data units. We introduce the idea of extra source or destination connectivity with respect to a certain S-T max-flow, and then we introduce two problems: namely, pre-cut protection and post-cut protection. Protection is implemented using the technique of network coding, and using coefficients from the binary field. Because of space limitations we only focus on the pre-cut protection problem. Specifically, we show that the pre-cut protection problem is NP-hard, we propose a heuristic approach to solve it, and we compare the performance of this heuristic to results from an optimal ILP. Simulations show that the performance of the heuristic is acceptable even on relatively large networks.