Sofiane Ouni

Real-time quality of service with delay guarantee in sensor networks

The sensor networks raise fundamental problems for the scientific community. These problems are due to the wireless communications, the density of nodes distribution, the resources constraints (energy, processor and memory), the low nodes reliability and the strongly distributed nature of the supported application. The specific particularities of the sensor networks make it difficult the guarantee of the real-time quality of service which remains currently an important challenge. This paper presents novel protocol solutions for sensor networks in hard real-time environments. We try to get the optimisation of both constraints at the same time: real-time communication delay guarantee and the energy consumption. Our routing and medium access strategy allows the load balancing, the interferences reduction and the real-time guarantee, for being based on a dynamic generation of real-time communication trees. The simulation results show that our routing and medium access protocols help the traffic load balancing to increase the system lifetime and the guarantee of real-time traffics.

DSR based Routing Algorithm with Delay Guarantee for Ad Hoc Networks

Applying real time applications with deadline guarantee for Ad Hoc network seems to be a difficult challenge and to be an impossibility in general. This is due to the features of such networks. However, if we make an assumption on node mobility, traffics and reliability with predictable features, we can get real time guarantee. That is the goal of our present work. In this paper, we present a new routing algorithm based on DSR to get the guarantee of Hard real Time traffics. This algorithm is based on a node resource reservation for each traffic path. An admission control is made to preserve the reserved node resources and is adapted to node mobility by preventing path breaking. Our routing protocol is implemented under NS2 and simulation results perform more stable time features than classical routing.

Association criteria to optimize energy consumption and latency for IEEE 802.15.4/ZigBee wireless sensor networks

Technology of wireless sensor networks has grown thanks to technological developments in different areas related to microelectronics. However, to become truly ubiquitous, these networks must overcome a number of challenges such as resource constraints (energy, memory, etc.), quality of service (bounded transmission delays), etc. It follows that the energy conservation and latency minimization are critical functions for sensor networks. The ZigBee standard, combined with IEEE 802.15.4 offers the features that best meet the needs of wireless sensor networks. This standard defines the association procedure that allows the sensors to self-organize into a topology through which data can be routed to their destination. In this paper, we propose new approaches of association designed to optimize the topology formation of IEEE 802.15.4/ZigBee network. Then, we use simulation to validate their efficiency in terms of energy saving and optimization of latency for IEEE 802.15.4/ZigBee sensor networks.

Cooperative Association/Re-association Approaches to Optimize Energy Consumption for Real-Time IEEE 802.15.4/ZigBee Wireless Sensor Networks

The major challenge of real-time Wireless Sensor Networks stills the optimization of both constraints: energy consumption, to get long network lifetime and the communication delay, to meet real-time requirements. In the context of IEEE 802.15.4/ZigBee networks, the association procedure has a direct effect on building paths optimizing those constraints. In this paper, we are interested on the definition of an ideal approach of load balancing to fairly distribute energy consumption among nodes in IEEE 802.15.4/ZigBee WSNs. This approach leads to conserve energy of each node in order to extend the network lifetime. To be closer to this ideal, we propose new dynamic association/re-association approaches allowing path alternation relative to association criteria and their threshold parameters. The implementation of those approaches in NS2 simulator highlights the efficiency of cooperative and dynamic association criteria particularly the one based on the sum of the inverses of remaining energy. Indeed, this approach gives better results with regard to energy distribution according to ideal approach which leads to a longer lifetime. It also performs lower latency for real time communication.

Energy consumption analysis to predict the lifetime of IEEE 802.15.4 wireless sensor networks

A Wireless Sensor Network (WSN) consists of several sensor nodes with limited energy source. Lifetime of such network depends on the residual energy of its nodes. However, to successfully achieve the designated mission, WSNs should meet a required lifetime. Accordingly, computing lifetime allows a better prediction of the network availability. Therefore, this paper proposes an analytical model to predict the lifetime of IEEE 802.15.4 wireless sensor networks. Our proposed model is based on a complete energy consumption analysis. In fact, it considers the most important sources of energy consumption such as overhearing, idle listening, overheads and collisions due to interference. Compared to simulation results and other analytical approaches, our model gives a reliable lifetime prediction for a given sensor network scenario.

Optimized TDMA multi-frequency scheduling access protocols for sensor networks

In order to eliminate the interferences between neighbors and reduce the delay, with the respect of the real time constraint, we focus our work on the multi-frequency scheduling access protocols. In this paper, we review the most multi-frequency MAC protocols which deploy more than one frequency to transmit data to neighbors. After that, we propose an optimized TDMA multi-frequency scheduling access contribution for sensor networks which is based on three phases. In the first phase, we assign frequencies based on parent node reception in communication tree structure. The reduction of the number of frequencies is provided, in phase two, based on the principle of the graph coloring. Finally, in the last phase, we reduce frequencies to the maximum number of allowed frequencies with TDMA (Time Division) clustering principle. Our results are compared with the other works and have showed that our strategy can remove more interference between neighbors, optimize communication delay and adapt with the maximum number of frequencies in WSN (Wireless Sensor Networks).

Data aggregation and pipelining scheduling protocols for real-time wireless sensor networks

Critical applications in real-time wireless sensor networks require bounded service latency and energy consumption optimisation. This motivates us to design a novel scheduling approach, which integrates communications with pipelining and data aggregation. The aim of our aggregation technique is to reduce the number of messages, which consequently reduces energy consumption. However, our pipelining technique is defined to minimise communication latency. Our approach is applied on tree wireless sensor networks and in a real-time context, which requires communications without interference. Thus, we use TDMA and a multi-frequency access method based on parent relationship clustering. In this context, we made the necessary mathematical formulation to determine communication response time and to predict the impact of real-time constraints, particularly deadlines, on communication delay. Our simulations outperform the efficiency of our approach to comprise between communication delay and energy consumption constraints.

End-to-End Delay Guarantee for TDMA-Based Ad Hoc Networks with RT-DSR Protocol

The hard real time guarantee in the Ad hoc networks seems to be a difficult task due to the features of such networks. For that reason, it is not sufficient to propose a routing protocol which guarantees the hard real time transfer. We are also interested in the medium access method. In this paper, after describing our routing protocol RT-DSR (Real-Time DSR), we present the TDMA (Time Division Multiple Access) access methods for the Ad hoc networks and we propose a TDMA approach which can offers better delay guarantees. The simulation results performed with NS2 simulator show that our protocol RT-DSR offers better delay values than the DSR routing protocol and prove that the TDMA access method is more suitable than the 802.11 access method.

A Fast Handover Procedure Based on Smart Association Decision for Mobile IEEE 802.15.4 Wireless Sensor Networks

Recent technological progress in microelectronics has allowed a considerable development of Wireless Sensor Networks. These networks are deployed in several relevant applications such as healthcare and wildlife which require the support of sensor nodes mobility. However, this mobility is a real threat in breaking communications and packet loss accordingly. This paper proposes a fast handover procedure based on a smart association decision to handle mobility and to ensure continuous communication in IEEE 802.15.4 WSNs. In our proposed procedure, the mobile node can anticipate the coordinator change upon detecting the degradation of the link quality indicator. Then, it performs a fast re-association based on our smart criterion and it resumes the forwarding of stored packets. Simulations show that our fast handover procedure ensures better network performances than a similar approach.

New Re-association Procedures for Reliable Handover in IEEE 802.15.4 Wireless Sensor Networks

Many Wireless Sensor Network applications such as healthcare and wildlife monitoring require the support of nodes mobility. However, the mobility has large impact on the network behavior namely the dynamic network topology and the synchronization loss of mobile nodes. Given that, this paper proposes two new re-association procedures for reliable handover to deal with nodes mobility in IEEE 802.15.4 WSNs. In the first approach, the handover process is triggered by parent nodes based on their residual energy. In the second approach, the handover process is triggered by mobile nodes upon location change. Both approaches allow to minimize data packet loss during the handover process while optimizing energy consumption. We used simulation in order to highlight the efficiency of our proposals namely in terms of network throughput and reliability.