M3: Multimedia for Mobile nodes in a wireless Mesh network. Collaborative project with Stanford University, USA, Funded by AEA $744,808, 2009 - 2010

• PI at KAUST: Basem Shihada
• PIs at Stanford: Philip Levis and Fouad Tobagi

This project is committed to address the key problems for supporting multimedia over wireless mesh networks.  We have identified that power, delay, and buffer management are the key challenges. We provided solutions to the above challenges by deriving novel algorithms, novel protocols, and optimizations. We have also evaluated our achievements over IEEE 802.11 (WiFi) wireless mesh testbed and extended our studies to the IEEE 802.16 (WiMax) research domain and testbed experiments.

Video Multicasting over WiMax, Collaborative project with University of Waterloo, Funded with $210,000, 2010 - present.

• PI at KAUST: Basem Shihada
• PIs at U. of Waterloo: Pin-Han Ho and James She
• PhD Student: Chun Pong Lau
• Msc Student: Ammar Meer

The work on multimedia over wireless mesh networks has inspired us to explore other wireless transmission standards. For this reasons, we have chosen IEEE 802.16 (WiMax). In this work a test bed experiment is designed to demonstrate the framework on the Cross-Layer Superposition Coded Multicast for Robust IPTV Services over WiMAX. The Goal of this experiment is to transmit scalable video by Super Position Coding (SPC) to demonstrate different video quality will be received under different wireless channel quality. If the channel quality is good enough, enhanced quality video can be received. If the receiver signal is not strong enough, at least the base quality video can be played. The Scalable Video is encoded by JSVM 9.18 developed by Heinrich Hertz Institute. The generated video is extracted into 2 layers, Base and Enhancement by BitStreamExtractor developed by the networking group at KAUST. The testbed consists of two emulators shown in the figure. Two PCs are connected to the emulators by LAN cable. The emulators are connected by RF cables through a variableattenuator from up converter on transmitter to down converter on receiver. We are also working on extending this work to tackle experimenting transmitting scalable video over different channel widths.

Buffer Management in Wireless Mesh Networks, 2010 - present

• PI at KAUST: Kamran Jamshaid

In this work, we analyzed the problem of buffer sizing for TCP flows in 802.11-based Wireless Mesh Networks. The objective of the work is to maintain high network utilization while providing low queuing delays. We observed that the problem is highly complex due to the time-varying capacity of the wireless channel as well as random access mechanism of 802.11 MAC protocol. While arbitrarily large buffers can maintain high network utilization, this comes at the cost of large queuing delays. We proposed a novel sizing routing buffers collectively for a set of nodes within mutual interference range called the ‘collision domain’. We provided a buffer size which is large enough to saturate the available capacity of the bottleneck collision domain that limits the carrying capacity of the network. This neighborhood buffer is distributed over multiple nodes that constitute the network bottleneck. A transmission by any of these nodes fully utilizes the available spectral resource for the duration of the transmission; thus sizing routing buffers collectively for this bottleneck allows us to have substantially smaller buffers (as low as 2−3 packets) at individual nodes without any significant loss in network utilization. We proposed heuristics to determine these buffer sizes in WMNs. Our results show that we can reduce the end-to-end delays by a factor of 10 at the cost of sacrificing only 5% of the network capacity achievable with large buffers.

Power and Delay Optimization for Data Transmission in Wireless Mesh Networks, 2010 - present

• PI at KAUST: Li Xia

In this work, we considered the optimization of transmission power and delay in a wireless mesh networks. Our target is to dynamically determine a set of transmission rates for all the nodes in the network according to the nodes’ data queue length and channel states, in order to initiate an optimal tradeoff between the power consumption and queuing time of transmitted data. We formulated this problem via a suite of modeling approaches, including the Jackson network model for data transmission and a Markov model for formulating the channel states transition. Performance difference equation is derived when the system parameters are changed arbitrarily. Some interesting properties about the optimal solutions are proved and a policy-based iteration algorithm is developed to find the optimal solution with an online implementation manner. This work is considered the first to tackle the power and delay optimization over multiple nodes in a wireless mesh networks. We have also simulated the algorithm over 4 to 6 hops and proved to achieve better data transmission. We anticipate that this work can also be applied to the ad-hoc networks where multi-hop transmissions exist.

Collaborate Inspection and Repair by a Fleet of Autonomous Underwater Vehicles, Collaborative project with KFUPM, First round approved, KACST Funding, 2011

• PI at KAUST: Basem Shihada
• PI at KFUPM: Sami El-Ferik
• PhD Student: Ahmad Showail
• MSc Student: Muhamad Felemban
• MSc Student: Matthew Debont

The project aims to develop an unmanned fully autonomous under water vehicles (UWV) system and to demonstrate its operation using collaborative and distributed control. We overcome the challenges of underwater communications by an integration of sensor and multi-hop networks. Human intervention is reduced. The vehicles are wirelessly coordinated to reliably transmit the control data in flow fashion. The vehicles are intended to work in relative independence from the mother vessel. In our solution model, we introduce a special class of sensor and mesh networks, called Flow Sensor-Mesh Networks (FSMN).

Cognitive Wireless Radio (CogWnet), Collaborative project with University of Michigan, 2010 - present

• PI at KAUST: Basem Shihada
• PI at U. of Michigan: Kang Shin
• PhD Student: Ismail ElQerm

CogWnet is a software framework to manage cognitive radio functionality and provide optimal parameters configuration for wireless channel. The Framework is based on cross-layer design to collect sensory information from all layers in TCP/IP stack along with application QoS requirements. The CogWnet layers are: Communication Layer, Decision making Layer, and Policy Layer

TCP over Optical Burst Switching Networks (TCPOBS), 2006 - 2009

The primary objective of the project is to develop a suite of interoperable strategies that can facilitate the user domain TCP window congestion control mechanisms to well cooperate with the intrinsic characteristics/behaviours of the OBS transmission. We expect that the results of performance analysis are considered heuristic for gaining better knowledge and deeper understanding on the TCP performance behaviour over both the conventional Internet and the OBS networks. Network Simulator NS-2 is used, where the library in Optical WDM Network Simulator OWNS has been significantly extended.

User Control LightPath Project Funded by CANARIE CRC, 2002 - 2004

Canarie (CRC) and Cisco Canada arehosting a research and developmentproject entitled "User controlled lightPaths". It aims to design solution architecture for customizable optical network resource provisioning for Grid applications. The objective of the project is to provide a software system that allows users (e.g. research institutions, government departments, hospitals) to own and control lightpaths, enabling rapid provisioning of network resources across multiple independent management domains.  By allowing users to advertise unused portions of their lightpaths for lease by others, the system will also increase the pool of available network resources beyond what can be offered by carriers. The proposed software system manages lightpath-related data and interacts with lightpath cross-connect devices in order to provide users with the following functionality 1) partition the available bandwidth in an owned lightpath and spawn child lightpaths, 2) advertise an owned lightpath for lease, 3) lease an advertised lightpath from another user for a limited period of time, 4) concatenate a series of lightpaths to establish an end-to-end lightpath spanning multiple independent management domains, and 5) manage cross-connections between owned lightpaths as well as peering with lightpaths of other users.

Mini Projects:

Conceptual and Concrete architecture of open source DBMS called LEAP

Conceptual and concrete architecture of Open Network Management System OpenNMS

An implementation and performance evaluation of XTR public key Cryptosystem over wireless links

e-approximate Algorithm for SLA Negotiation

Architectural Description Languages and their Role in Component Based Design

Simulated Dijkstra's algorithm for shortest paths calculation (OSPF).

Implemented Parallel Sorting with Regular Sampling (Shi & Schaeffer), with Message Passing Interface on LAM using a maximum of 4 processes under one sequential processor.