A Smart Physical Layer for Highly Reconfigurable Mobile Networks

Cofin 2005 Massa DIT-PRJ-05-142

Status NOT active project
DISI role Coordinator
Project type Research Project
Dimension International
Acquisition date 2005-12-30
Start date 2006-01-30
End date 2008-01-30

Project details

Project astract The proposing research group is composed by three units: the Department of Information and Communication Technology of Trento (DIT), the Department of Biophysical and Electronic Engineering of Genova (DIBE), the Department of Electric and Information Engineering of Salerno (DIIIE). The group collaborates with the FIAT Research Centre (CRF), the FRACARRO Radioindustrie S.p.A and the Microwave Instrumentation Technologies LLC.<br/>The two-years research project has the main scope of studying, designing and developing a SW/HW integrated system for realizing the distributed intelligence at different levels of a mobile network with high re-configurability.<br/>The project will be developed with reference to a vehicular context considering interactions between mobile-nodes/mobile-nodes and mobile-nodes/infrastructures.<br/>The research activity will be developed in different workpackages (the definition of the system requirements and of the sub-systems specifications, the model of the channel and of the simulation environment, the design and realization of the sub-systems, the development of prototypes and demonstrations, the testing of the system in a real controlled environment) in order to realize the following topics strongly interconnected: the model of the transmission channel, the localization of the network nodes, the definition of adaptive transmission modalities, the definition of the smart antenna systems, and the adaptive routing.<br/>Products of the research will be the reports describing the functional and qualitative requirements of the system/sub-systems, the characterization of the propagating channel, the design of the different antenna systems (mobile nodes and infrastructures) with their control algorithms, the functional verifications and the standard conformity. Moreover, a final report will describe the results of a detailed experimental testing in a real controlled environment in order to validate the developed algorithms, the synthesis of the antenna system and the interactions with other elements of the system. Products of the research will also be the realization of prototypes of the main SW/HW components: the antenna system, the computer codes for the localization, and for the definition of the transmission and network modalities. Moreover, simulation environments for the analysis of the radiating systems will be realized.<br/>A workshop is also planned for presenting the results.
Keywords Smart Antennas, Electromagnetic Wave Propagation in Complex Media, SDR/SCR, TRANSMISSION MODE IDENTIFICATION ; DISTRIBUTED DETECTION ; LOCALIZATION ; ADAPTIVE ROUTING ; WLAN
Fundings 187223 €
Partners
  • University of Salerno
  • University of Genova
  • DIT - UniTN

DISI Sub-project details

Project astract The potentialities of wireless networks allowed developing a large amount of applications in several working environments. Among them, indoor applications played a major role: here, the main advantage is the possibility of establishing an IP connection without the need of cabling and complicated hw/sw procedures. In this framework the standards of the IEEE 802.11 family have been elected as a viable and low cost solution for short range communications able to ensure an adequate throughput both for data transmission and for delivery of multimedia contents. An applications that is gaining increasing attention concerns outdoor connectivity to nomadic users whose mobility can space from a restricted environment to more complex scenarios such as vehicular communications.<br/>In this scenario the network access is helpful in a large variety of situations, allowing data exchange and different kind of services for specific or general purposes. A possible application could be the vehicle-to-vehicle or vehicle-to-infrastructure data exchange in risk situations such as traffic, bad weather conditions, road works, accidents. Nevertheless, they present some limitations in extreme working conditions, specifically when mobility and re-configurability of the network are a desired constraint. These limitations have to be carefully evaluated.<br/>Both QoS requirements and resource limitations lead to the definition of adaptive solutions, in order to enable an optimal distribution of the resources introducing priority mechanisms. The introduction of some kind ofintelligencein the low levels of the network and a strong cross-layer adaptation may lead to a significant improvement of the system performance. In this context several important factors arise, such as the study of propagation phenomena in complex environments, the study of the radio channel and the multi-user interference, the development of real-time antenna systems able to optimize the communication in presence of interference/noise and the definition of adequate routing strategies. In order to reach the mentioned objectives it is necessary to study and define new integrated intelligence mechanisms that involve both the antenna system (physical layer) and the data routing techniques (network layer) on the basis of the channel behaviour knowledge. <br/>As far as the modelling of the radio channel is concerned, it is well known that its impulse response is strongly related to the propagation environment. Therefore, its representation through the formalism of Maxwell equations requires an accurate and detailed description of the propagation scenario, and this is often very difficult to obtain, both for time and cost reasons. If we refer to a strong time-varying scenario as for example the vehicular one or, more in general, where a high re-configurability of the network is needed, it is necessary to use non-deterministic or simplified models. These can be defined starting from a reduced set of geometric parameters that are able to provide a global picture of the transmission channel.<br/>Recently, new stochastic approaches has been proposed in order to study the electromagnetic propagation in outdoor environments. This method is based on the concept of random walks which is characterized by a reduced computational cost and by the possibility of extending the obtained results to other environments that can be considered as statistically equivalent.<br/>As far as the antenna intelligence is concerned, the Smart Antennas (SA) are receiving a growing interest because they are able to increase the performances of the wireless system. They can partially overcome the problems related to the increment of the necessary frequency bandwidths and of the number of connected users. The systems employing the features allow to increment the capacity, the spectral efficiency and the BER, widening the coverage area. The use of intelligent antenna systems can reduce the effects of the multipath fading and of the co-channel interference, allowing a reduction of the system complexity and the out-of-service probability by limiting the number of fixed base stations. Accordingly, the main characteristic of a SA system is the possibility of deleting the co-channel interference by focalizing the beam in the direction of the desired signal. The placing of the nulls or high attenuations towards the directions of arrival of the interfering signals, allowing the maximization of the Signal-to-Noise-plus-Interference-Ratio (SINR).<br/>A SA employs reconfigurable antennas arrays, in which the beam pattern can be shaped and directed towards a determined direction by varying the feeds of the single radiators. Towards this aim, suitable signal-processing algorithms are usually utilized. They are able to distinguish in real time the desired signal from the interferences detecting its direction of arrival (DOA). By so doing, the system is able to track the signal of interest and attenuate the interferences opportunely changing the feeds of the radiating elements. <br/>Some examples can be found in literature, which deal with adaptive antenna arrays (smart antennas) able to modify their beam pattern according to changes of the environment, both due to the presence of mobile stations or to interference signals (usually time-variant). Several optimization techniques have been developed, focusing the attention on two main problems. The former concerns the estimation of the direction of arrival (DOA) of the signals. Both approaches based on maximum likelihood algorithms or derived from them and on learning-by-examples techniques have been proposed to this aim. The latter concerns the real-time adaptive beam-forming, necessary to locate maxima of the beam pattern in correspondence to the DOAs (i.e, towards mobile stations or access points) and zeros in correspondence of interference signals. In this framework, two main approaches should be mentioned: the one by Applebaum and the one by Widrow. These methods theoretically permit to obtain very good performances in the adaptive beam-forming. Nonetheless, they are not able to treat discrete parameters. From an operative point of view, this implies the use of analogical phase shifters, thus leading to expensive (or even unrealizable) control systems.<br/>In order to overcome these drawbacks, new optimization techniques have been proposed for digital adaptive array design. Among them, approaches based on genetic algorithms seem to be very promising. On these approaches and on their application on the design of ad-hoc antennas for WLAN network will focus a part of the research project. Particular attention will be devoted to the development of suitable GA-based optimization techniques able to use the memory about the temporal evolution of the electromagnetic environment in order to obtain a real-time adaptive beam-forming.<br/>As far as the upper layers are concerned, one of the most critical aspects of the envisaged application is related to the adaptation of medium access and routing protocols, required to manage the high re-configurability of network nodes. On the other hand, the range of applications of such networks is extremely broad and interesting, making it worth the study of efficient protocols to effectively solve the above problems. In particular, the rapid and continuous modification of the network topology may also have positive effects. In fact, although it determines the need for more efficient routing algorithms, it has been recently demonstrated that it can also produce an increase of the network capacity. More specifically, it is possible to achieve an increased throughput whenever the directions of the movement are limited. This characteristic is typical, for instance, of automotive applications. In this sense, one can think to develop some smart routing techniques able to exploit this property in order to better utilize the available resources. In particular, exploiting the knowledge of relative or absolute node localization data can improve efficiency and QoS. Some interesting results in this framework have been published in some recent works on ad-hoc network routing. Further improvements can be obtained using the knowledge about motion direction to create stable paths in high mobility networks.<br/>Finally, it is worth considering the transmission over the considered networks acts in multi-hop mode. The communication involves therefore a pair of mobile nodes or a mobile node and the infrastructure. In both case, the cross-layer interaction between the two entities should be managed in a smart way. It is to be mentioned that recent works in this field show that the use of suitable cross-layer optimization approaches between physical (smart antenna), MAC and network (routing) levels may boost the performance of ad-hoc networks with high density of nodes, by allowing the creation of clusters of nodes along the preferred communication direction.
Keywords SMART ANTENNAS ; ELECTROMAGNETIC WAVE PROPAGATION IN COMPLEX MEDIA ; OPTIMIZATION TECHNIQUES ; WLAN
Fundings 85146 €
Manager Andrea Massa