MY RESEARCH: OVERVIEW ON MY MAIN ACTIVITIES AT DIMES – UNICAL (ITALY)
Since 2004, I am focusing my research activity on some important issues regarding a wide variety of telecommunications areas: from Ultra Wide Band systems, to WirelessLAN environments, Satellite Communications, Ad-Hoc Networking, Vehicular Ad-Hoc Networking, Quality of Service Protocols, Channel modeling, Broadband communications systems like WiMAX, Swarm intelligence, Artificial Intelligence, Neural networks, Mobility patterns analysis and prediction, Power Lines Communication systems. Recently I started to investigate some issues and applications about Wireless Sensor Networks and Electric Vehicles. See my publications page (PUB) for more details about the results of my work.
1. WirelessLAN and Channel Modeling
Nowadays, wireless communication systems are becoming popular for their ability to satisfy emerging users’ needs. A wireless system should ensure users a certain level of Quality of Service (QoS), so new protocols must be often investigated and proposed. Avoiding service disruptions is a suitable condition for wireless environments and an appropriate bandwidth managing policy should be employed: so dynamic bandwidth schemes are investigated and proposed by choosing the appropriate level of resources to be assigned to each user admitted into the system. In addition, wireless communications are characterized by signal level fluctuations and it is very important to model wireless channel behavior by the introduction of the stochastic theory. Predicting users movement into the coverage system has also its relevance, so statistical approaches to analyze users’ movements are proposed, in order to make some predictions on users’ future patterns. Here you can find some of my publications about this topic: [C001], [C002], [C003], [C004], [C006], [C011]; for more details, newer articles and journals, please refer to (PUB).
2. Ultra Wide Band Communication Systems (UWB)
Ultra-wideband (UWB) radio communicates with baseband signal pulses of very short duration (typically the duration is few nanoseconds), with very low power spectral densities values (typically few microW per MHz), and their energy is spreaded on a very large band, so they can coexist with incumbent systems in the same frequency range. These characteristics and the possibility of achieving sufficiently higher data rates without the needing of increasing transmitter power make UWB technology a possible candidate for short (< 20 meters) and medium (< 100 meters) range multiple access communications in dense multipath indoor and outdoor environments. The classic channel modeling approach models the phenomena of wireless channel through an impulse response and so it can work only at the physical level, not suitable for working at higher levels. This is not a trivial problem, especially if an investigation on the higher levels of ISO/OSI architecture is needed: in fact, in this case, a model describing the trend of wrong packets in time (hence not the signal corruption at the physical layer) is needed. For this purpose, there are many approaches and efforts to obtain an accurate high level channel model based on Discrete Time Markov Chains (DTMCs – useful in every simulation context). In particular, main studies regard the concept of error trace analysis and degradation level in different time observation windows. In the literature, many works have approached problems related to physical and MAC layers of UWB system, and many others regard the UWB routing issues. Traditional ad-hoc network approaches, based for example on minimum hop count or geometric criterions, can be inadequate for these network because UWB systems are strongly affected by the mutual interference between nodes: that is the most undesired problem in UWB wireless networks, because it can cause an irretrievable level of degradation. Therefore a routing protocol, that does not take into account interference between nodes, could lead to choose a wrong path, in terms of signal degradation: the distance between the source and destination can be minimized, but the level of interference may be too high if new metrics are not defined. There is a lot of research effort concerned on the definition of new protocols, based on the concept of interference-aware (the choice of the optimum route could be made on the basis of interference perceived by the nodes). Here you can find some of my publications about this topic: [C012], [C013], [C016], [C019]; for more details, newer articles and journals, please refer to (PUB).
3. Satellite Communications Systems
The management of services diversity within a high error rate, high-delay and limited wireless resource requires a careful and efficient use of bandwidth, capable of meeting the needs of individual users. The scientific community is trying to give greater emphasis to existing Geostationary Earth Orbit (GEO) satellites that have good characteristics in terms of coverage area and do not have the disadvantage of mobility, because they are similar to a fixed point in the sky. These satellite platforms have, however, the disadvantage of high delay and high interference. In particular, the attention is focusing on the integration of these new satellite networks with fixed structures (wired), in order to provide users with an ever greater bandwidth and service, always active in any part of the globe, but also with new platforms called High Altitude Platforms (HAPs). HAPs have many advantages, such as propagation delay (lower than the classical platform GEO), ease of installation and, therefore, the possibility of setting up telecommunication networks in impervious areas where, otherwise, it would be very difficult to make a wired network. The integration of these networks arises important issues concerning the management of Quality of Service (QoS) and the efficient routing within the GEO satellites-HAPs-terrestrial network. In particular, in this context, hybrid multicast algorithms based on multiple metrics (e.g., cost, bandwidth or delay) by using optimization criteria are interesting to analyze. The performed studies concern many research lines: – Architectures for Quality of Service (QoS) in networks is a hybrid-type satellite with particular reference to the integrated and differentiated services; – Management of traffic differentiation on the satellite platform; – Planning and implementation of scheduling algorithms for the management of different classes of traffic; – Planning and implementation of intelligent Call Admission Control (CAC) algorithms in DVB-RCS platforms; – Study of integrated hierarchical multi-layer platform with particular attention to layers 3 and 4 of ISO/OSI model. Here you can find some of my publications about this topic: [C007], [C015], [C022], [C037]; for more details, newer articles and journals, please refer to (PUB).
4. Vehicular Ad-hoc NETworks (VANETs)
In general, VANET networks represent a sub-category of ad-hoc mobile networks, i.e. telecommunications networks in which nodes perform their functionalities (such as routing and transport) in ad-hoc mode and with a certain degree of mobility. Devices that belong to VANET do not necessarily use pre-existing network infrastructure. Since the network topology is highly variable in time, it is necessary to frequently update the information that each node has about neighboring nodes, so the routing protocol (which also solves the problem) must be designed carefully. There are many studies and proposals in literature regarding the “theoretical” resolution of some issues related to VANETs as: a) In-depth analysis of the PHYsical and MAC layers of the IEEE802.11p standard, in order to understand how to take the advantage of the potential of devices that may be built-in on the vehicle (WAVE/DSRC transmitters), optimizing the transmission and reception of data in terms of power consumption and collisions on the transmission mean; b) In-depth analysis of the NETwork layer, in order to study how to forward the information, through which mobile nodes can communicate, for example, with a remote server that is responsible for carry out the route optimization; further investigation should be made about the type of routing to be used (reactive, proactive, geographical, positional, etc.), in order to obtain the best results. Here you can find some of my publications about this topic: [C032], [C033], [C034], [C039]; for more details, newer articles and journals, please refer to (PUB).
NEWER RESEARCH ACTIVITIES
5. Multicast Routing for Satellite and VANET communications
Multicast routing can be applied to different application types. For example it is well suitable for multimedia applications as video conference, video on demand and so on. Multicast is utilized in order to obtain a better allocation of the bandwidth and to distribute in a better manner the packets flow. In the unicast packet distribution there exists a path that distributes data for each connection between a source and a destination, therefore more bandwidth and more packets are sent on the network. Multicast permits to allocate a single path and it can distribute the entire data between one or more sources, towards the multicast destinations. This is possible thanks to multicast spanning trees, that are found on the network using specific algorithms. Many efforts have been given to the field of QoS multicast, in particular when more than one QoS metric is considered: the main issue is to find the best multicast tree in the network. In this scenario, simple algorithms cannot be used, owing to the problem complexity, that is known to be NP-complete; therefore, it is necessary to find the multicast tree in a reasonable time. In general, heuristic algorithms can be utilized. With these techniques is possible to consider different QoS metrics as end-to-end delay, jitter delay, packet error rate, bandwidth and so on. In this way different application requirements are satisfied and each multicast tree satisfies the specific QoS requirements. The found solution could be too far from the optimum one, if the algorithm is not well tuned. For this reason, new algorithms, that are auto-configurable in order to respond at the network dynamics, need to be investigated. New protocols, that deploy the aforementioned QoS-oriented algorithms, need to be investigated, considering the different characteristics that each network offer: satellite networks offer higher bandwidth but also higher propagation delay, HAP networks can offer a lower bandwidth than the satellite one but also lower propagation delay; a good trade-off between the different characteristics permits to obtain best performances and a better respect of the QoS constraints. Regarding vehicular networking, routing operations need always to be optimized: the possibility of considering composite metrics that account for the main network parameters is the subject of newer research activity, aimed at investigating the possibility of applying the multicast concept to vehicular environments.
6. Unconventional Prediction Approaches
Regarding the research activity no. 1, it is very important to employ an accurate mobility predictor, able to reduce errors when mobile hosts move in an infrastructured network. Different approaches are present in literature and the main ones regard Markov theory, Neural Networks and Data Mining (many others are present but do not offer good results in terms of accuracy). Additional research activity is mandatory, in order to investigate the convenience in employing a distributed approach instead of a classical centralized one, with the main aim of reducing prediction errors. Markov Chains or Neural Networks need to be structured and trained in a proper way: another open issue regards the representation of mobility data and the way it is considered during training phases. Here you can find some of my publications about this topic: [C041], [C045], [C051]; for more details, newer articles and journals, please refer to (PUB).