When Services Meet Security

Wireless Mesh Networks (WMN) is a promising technology due to its low cost and scalability properties. One important application of WMNs is namely the provision of services in a uniform manner. For this service provision the concept of Service Oriented Architecture (SOA) and hence Versatile Service Oriented WMN (VESO-MESH) arises. This kind of technology makes the network also vulnerable to attacks if no security measures are taken. In this work the security issue in a VESO-MESH is addressed. Hence an innovative key distribution and encryption mechanisms are shown. Furthermore a cryptographic analysis is presented.

Naturally Inspired Multi-model Context Recognition

Ubiquitous computing is an emergent field on which information processing is integrated into daily tasks in the most natural way possible. One of the most popular and complex application might be a Smart Home, on which users can interact with their home environment in order to achieve certain interests. These interests may include: preservation or improvement of quality of life, security, independency, etc.

On this project we focus on the security aspects of a smart home by establishing a context mapping between users and available functions. That is each user has its own context, only users with enough privileges may access a given functionality at a given time. For example intruders shouldn’t have any access at all, guests should have limited access, and children shouldn’t be able use kitchen functions which may become a hazard for the house (e.g. fire).

For us humans this is relatively simple since we have enough intelligence to distinguish between who we know and who we don’t. For the case of computers there has been work done to identify people by using only some classifier based on a certain input type (e.g. image, audio).

We propose to implement a multi-model classifier using data from images and audio samples and fuse them together. Our goal is to improve the accuracy and reliability of current literature projects. First we will make a simple one class problem (i.e. is it X or not) with pre-processing of input to have a basis of how well we can perform. Then we will complicate things by transforming the problem into a multi-class (i.e. does it belongs to known people and who is it).

Wireless Mesh Networks for Smart Homes

Wireless mesh network (WMN) have proven to solve many issues of the current world, using cheap solutions. Their broadcast nature makes them an attractive solution to provide services effectively. Currently, hundreds of municipalities are implementing or planning to implement meshed Wi-Fi for public Internet access, public safety, and business communications.

Although many applications are based on wide area coverage one cannot limit their capabilities to wide open spaces. In this project we encounter the challenge of developing a WMN for a small, but full of interference, area. Moreover, we propose to design a prototype for a smart home using a WMN as backbone for the nodes and Xbees to communicate devices in the home. The Texas Instruments microcontroller MSP430 will be used to control the devices remotely at a single location in the house.

This project consists of the design and implementation of a smart home prototype. We will use a MSP430 microcontroller to centralize the logic of the home along with various xbees to communicate wirelessly with all the devices. The house will have several sensors installed to detect the presence of either audio, movement, absence of people, etc. The lights will be activated with voice command.

Local Service Network

Wireless mesh network (WMN) is a promising technology for the next-generation wireless communication systems because it can support broadband services with ubiquitous coverage and with relatively low cost. Currently, hundreds of municipalities are implementing or planning to implement meshed Wi-Fi for public Internet access, public safety, and business communications.

Due to their broadcast nature, WMNs can provide services in a highly effective manner. Clients requesting services can connect to the mesh using the closest mesh router (MR), and obtain the services with better performance than that of a far network. In this project we develop a system analog to that of Global Positioning System (GPS), namely Local Service Network (LSN). Moreover, we compare the availability of services and performance of the system in contrast to the traditional satellite version.

This project consists of the design and implementation of a Local Service Network (LSN). Just like a GPS, the LPN promises to have knowledge of the surroundings providing information on maps and points of interest. We will create a wireless mesh network to provide the services of the maps and points of interests around the area, using the Linksys WRT54GL routers. A database will be stored inside of storage devices connected through Ethernet to the routers and accessed using wireless clients and location positioning services like Skyhook.

Service-Oriented Routing Algorithm (SORA)

Wireless mesh network (WMN) is a promising technology that can provide cost-effective solutions to cover a rather large area. Despite this important feature, the future generation of WMNs needs to support more and more applications, such as voice over internet protocol (VoIP) and multimedia content distribution. To efficiently support these diverse demands and to effectively utilize the wireless network, service-oriented network layer design has been proposed recently. A key idea of such a service-oriented design is that a user only needs to specify the service to the network, instead of the destination address in a traditional manner.

In this project, we present a simple solution, namely, the service-oriented routing algorithm (SORA), to enable service orientation in WMNs. The proposed scheme works in a similar manner as the Domain Name System (DNS). Specifically, SORA associates IP addresses to services just like the DNS associates IP addresses to domain names. SORA calculates the best route using the number of hops from a source to an arbitrary destination. This destination is chosen to be the one closest to the source and that has the service that it requests.

Our experiments show that the proposed algorithm can efficiently provide distributed services in wireless mesh networks.

Channel Selection and Network Coding (CHANET) Protocol

In recent years, IEEE 802.11 based wireless local area network (WLANs) have been widely deployed and the cost of IEEE 802.11-enabled routers a nd adapters are very low. To effectively establish wireless networks in various scenarios, it is important to exploit such inexpensive commercial off-the-shelf (COTS) products.

In this project, we address this issue and propose a novel medium access control (MAC) scheme based on the idea of network coding and intelligent channel selection. Specifically, we will design a new sub-layer, named CHANET, between the network layer and the legacy IEEE 802.11 MAC layer. In this manner, the cost for updating the system can be minimized, while the performance can be improved efficiently.

The first technique discussed is intelligent channel selection using Fuzzy Logic; with it the mesh can be transformed into a full-duplex connection, meaning that can receive and transmit at the same time. Of course that in order to achieve this the nodes need to have at least two radios, if this is not the case it can still send the information using the channel that is being used the less.

The second technique involves network coding at the network layer. Using DSP algorithms, the nodes in the mesh can create linearly independent equations that can mix N unknown packet to be decoded at the destination. This allows more data to be sent through the mesh utilizing even better the channel capacity, using multicast messages. The encoding of the packets is done while the system is scanning for the best channel to be use to transmit.

Our simulation results show that the proposed scheme can substantially improve the throughput and delay performance of the wireless network.