Hamdy S. Soliman

A neural networks approach to image data compression

We present a novel neural model for image compression called the direct classification (DC) model. The DC is a hybrid between a subset of the self-organizing Kohonen (SOK) model and the adaptive resonance theory (ART) model. The DC is a fast and efficient neural classification engine. The DC training utilizes the accuracy of the winner-takes-all feature of the SOK model and the elasticity/speed of the ART1 model. The DC engine has experimentally achieved much better results than the state-of-the-art peer image compression techniques (e.g., JPEG2000 and DjVu wavelet technology) especially in the domains of colored documents and still satellite images. We include a comprehensive analysis of the most important parameters of our DC system and their effects on system performance.

A smart forest-fire early detection sensory system: Another approach of utilizing wireless sensor and neural networks

In this paper, we analyze the potential of combining wireless sensor networks with artificial neural networks (ANNs) to build a “smart forest-fire early detection sensory system” (SFFEDSS). We outline our new SFFEDS system in which temperature, light and smoke data from low-cost sensor nodes spread out on the forest bed is aggregated into information. This information is spatially and temporally labeled into knowledge which will be encoded as input to ANN models that convert it into intelligence. At the top tier of our system, the trained neural models make intelligent decisions and report fire in its early stages based on gathered field knowledge. In our experimentation, we extended the sensing capability of the MicaZ sensor motes by attaching external smoke detectors of our own design. The results are very promising as the SFFEDSS unit is able to not only detect fire but also accurately report the direction of fire progress which is deduced from the wind direction.

An Efficient Routing Approach over Mobile Wireless Ad-Hoc Sensor Networks

In this paper, we efficiently adapt the prominent Ad Hoc On-Demand Distance Vector (AODV) routing protocol with a reactive Local Link Repair, AODV-LR, for effective deployment in restricted power-budget and bandwidth mobile ad hoc sensor networks (MASNET). We introduce a better replacement mechanism to the local repair phase of the AODV. Our new approach is a preemptive, self-repairing AODV (called AODV-PSR) scheme that is able to find an alternative link to a failing link. The new preemptive protocol achieves better performance than the ADOV-LR, due to its avoidance of packet buffering delay and excess use of control messages during link repair. Experimental results show a much lower obtained packet delay, a higher packet delivery ratio, and lower control message overhead. Hence, our mechanism is amenable to deployment in areas of restricted power-budget and bandwidth, such as a MASNET domain.

Enhancing AODV routing protocol over mobile ad hoc sensor networks

In this paper, we efficiently adapt the prominent Ad hoc On-demand Distance Vector (AODV) routing protocol with a reactive Local Link Repair (AODV-LR) for effective deployment in Mobile Ad hoc Sensor Networks (MASNETs) with restricted power and bandwidth. We introduce two replacement mechanisms to the local repair phase of the AODV. Our new approaches are self-repairing AODV algorithms called Reactive Self-repairing AODV (AODV-RSR) and Pre-emptive Self-repairing AODV (AODV-PSR). These two algorithms are able to find an alternative route to a failing link without depending only on broadcasting Route Request (RREQ) packets. Experimental results show that the new algorithms achieve better performance than the AODV-LR and Self-repair Algorithm (SRAODV) by obtaining lower packet delay, higher packet delivery ratio and lower control message overhead.

Efficient geographic routeless routing protocols with enhanced location update mechanism

In this paper, we introduce and evaluate two efficient controlled routeless routing protocols: Static and Dynamic Channel Width. They combine Location-Aided Routing (LAR) and Received signal strength-Aided Flooding (RAF) protocols for effective deployment in the very challenging domain of Mobile Ad hoc Sensor Networks (MASNETs). Our new approach maintains different controlled routeless geographic routing protocols that construct a virtual broadcasting channel between source nodes and their destinations. Hence with no predefined source-destination route, i.e. routeless, only some of the sensor nodes inside the channel are allowed to contribute in forwarding packets. Promising simulation results show that our protocols are able to outperform RAF, LAR, and Beacon-less routing protocols with respect to energy consumption, packet delivery ratio and end-to-end delay.

Application of synchronous dynamic encryption system in mobile wireless domains

Motivated by the tradeoff between security and efficiency performance parameters that has been imposed on all modern wireless security protocols, we designed a novel security system that gained in both parameters. Our system is based on stream ciphers for their speed, but maintaining a much more solid and proven security. Such security strength stems from the novel deployment of permutation vectors and the data records in the regeneration of the secret key. Moreover, the involvement of the former results in an adaptive and efficient data integrity mechanism that relies on error propagations in the data stream. Simulation results show that our security protocol is much faster than peer mechanisms such as WEP and CCMP. Hence, we anticipate a great opportunity to deploy our system in environments with scarce bandwidth, which are the most vulnerable; specifically the wireless domain.

Detecting border intrusion using wireless sensor network and artificial neural network

Monitoring movement across national borders is a challenging problem due to several economic and technical issues. Due to the vast size, remoteness and other geographical confines of border regions, technical solutions are necessary to complement the limitations of manpower. In this paper we discuss our research in developing a system for detecting border intrusion activity by combining wireless sensor networks with artificial neural networks (ANNs). The key idea is to use ANN models to discover distinct patterns that describe an intrusion activity and use these patterns to train the ANN model which will then recognize intrusions and other abnormalities. We present a border intrusion detection system in which light and sound data from low-cost sensor motes spread out on the field is used to help ANNs make automated decisions and report intrusion activity. Our experimental results show that our border intrusion detection system can be used to monitor the borders without constant human supervision1.

Mobile Cloud Gaming: Issues and Challenges

Recent developments in mobile, cloud, and graphics processing technologies have enabled mobile cloud gaming, a gaming model where players use mobile devices to play graphics-intensive games that run remotely on cloud servers. This delivery paradigm is called Gaming as a Service GaaS. GaaS is used to stream computer games across the Internet. It gives rise to various technical, legal, and ethical issues. In this paper, we present the current state of the art in GaaS along with open issues and research challenges.

Application of Synchronous Dynamic Encryption System (SDES) in Wireless Sensor Networks

Inherent to the wireless sensor networks are the two major problems of the broadcasting vulnerability, the limited computational capability and power budget. Even though security is a must in most applications, current sophisticated security protocols are not amenable to the primitiveness of the sensors. In this paper, we introduce a novel security protocol for wireless network of sensors that is very secure, yet simple and efficient. At the core of our security protocol is a simple and fast stream cipher cryptosystem that utilizes permutation vectors as encryption keys, forcing an intruder to a brute-force time complexity of Ω(2^n). In addition, our mechanism alleviates the effect of sensor capture, via its synchronized re-keying feature. In addition to the encryption efficiency, our system utilizes the group deployment of newly joining sensors for sensors power budgeting considerations. Experimental results show very promising future of our system in the wireless networks domain, excelling over other peers of modern cryptosystems (AES, DES, TripleDES), especially in the power budget arena.

Efficient Controlled Routeless Routing Protocol With Fixed Channel's Width for Mobile Ad Hoc Sensor Networks

In this paper, we introduce an efficient Controlled Routeless Routing with Fixed Channel Width protocol (CRRFCW) that combines LAR and RAF protocols for effective deployment in restricted power-budget and bandwidth mobile ad hoc sensor networks (MASNETs). Our protocol is a controlled routeless geographic routing protocol that constructs a virtual broadcasting channel between sensor nodes and their sinks. Only some of the sensor nodes inside the channel are allowed to contribute in forwarding packets. The simulation results show that our protocol outperforms both RAF and LAR protocols in number of retransmission, energy consumption, end-to-end delay and packet delivery ratio. Hence, our protocol is amenable to deployment in areas of restricted power-budget and bandwidth, such as the MASNET domain.