Papers by Christian Roland
HAL (Le Centre pour la Communication Scientifique Directe), Apr 1, 2009
Notre application s'inscrit dans un contexte d'analyse de signaux radar. Notre objectif e... more Notre application s'inscrit dans un contexte d'analyse de signaux radar. Notre objectif est de concevoir un système permettant d'effectuer en temps réel une détection d'énergie en temps et en fréquence sur un signal large bande. Le CAN du système échantillonne le signal en fréquence intermédiaire à une fréquence de 1,33 GHz. La bande utile du signal est de 500 MHz. Les échantillons sont ensuites regroupés par paquet de 8 échantillons consécutifs et sont entrées dans le FPGA à une fréquence d'horloge de 167 MHz. Les échantillons sont codés sur 8 bits. Une première étape de démodulation numérique (non décrite dans ce résumé), permet de démoduler le signal et de générer les échantillons complexes en bande de base (4 échantillons par cycles d'horloge). L'étape suivante consiste à effectuer l'analyse temps fréquence en temps réel. Le principe est de diviser la bande du signal en N = 64 sous-bandes et de mesure l'énergie reçue dans chacune de ces sous-b...
2019 IEEE Computer Society Annual Symposium on VLSI (ISVLSI), 2019
2018 Twelfth IEEE/ACM International Symposium on Networks-on-Chip (NOCS), 2018
2018 Twelfth IEEE/ACM International Symposium on Networks-on-Chip (NOCS), 2018
2019 16th International Symposium on Wireless Communication Systems (ISWCS), 2019
The use of Unmanned Aerial Vehicles (UAVs) is rising constantly whether for leisure or profession... more The use of Unmanned Aerial Vehicles (UAVs) is rising constantly whether for leisure or professional purposes in civilian or Defense domains. We consider in this study small civilian aerial drones of different types, which are low cost, available off the shelf and so affordable for individuals. Simultaneously, they have also raised security concerns for critical sites such as nuclear stations, strategic locations like official buildings, crowded places as stadiums, etc. The aim of this paper is to provide a survey of the risks assessment with and for UAVs in general. Regarding the security concern we pay a specific attention to attacks that are facilitated and can benefit from an easy access to Software Defined Radio (SDR) boards that can be embedded in the UAV or in the ground segment.
Proceedings of the 13th IEEE/ACM International Symposium on Networks-on-Chip - NOCS '19, 2019
Frequenz, 2004
This paper discusses the implementation of modulation chains for multi-standards communication on... more This paper discusses the implementation of modulation chains for multi-standards communication on a dynamically and partially reconfigurable heterogeneous platform. Implementation results highlight the benefit of considering a DSP/FPGA platform instead of a multi-DSP platform since the FPGA supports efficiently intensive computation components which reduces the DSP load. Furthermore partial dynamic reconfiguration increases overall performances compared to total dynamic reconfiguration since there is 45% of bitstream size reduction which leads to a 45% decrease of the whole reconfiguration time. The implementation of modulation chains for multi-standards communication proves the availability of new technology to support efficiently Software Defined Radio.
For providing every service, every where on every network, the solution currently studied in the ... more For providing every service, every where on every network, the solution currently studied in the telecommunication community is to propose re-configurable terminals. This is now possible thanks to Software Radio technology. The general tendency is to make these terminals sufficiently « intelligent » so that they themselves choose their own function mode. The technique proposed in this article allows this result to be arrived at via automatic research carried out by the receiver without any user intervention. A first study of the parameters of all the current systems has allowed us to identify the « channel bandwidth » as a discriminating parameter. We therefore propose to recognize this parameter using Radial Basis Functions Neural Networks. We have defined a new error which combines two others known errors function in order to obtain convincing results.
In many efficient wireless communication systems, the demodulation tasks resume to solve a combin... more In many efficient wireless communication systems, the demodulation tasks resume to solve a combinatorial optimization problem of the form y = Hx+b where H is a real m×n channel matrix, b is a noise vector, y is the received vector and x is a discrete ndimensional vector of transmitted information. The Maximum Likelihood (ML) detection of y = Hx+b is known to be a NP-hard problem in the general case. In this paper, a new sub-optimal algorithm for ML detection problem is proposed. This algorithm called HISD (Hyperplane Intersection and Selection Detector) is based on a geometrical approach. Comparing to existing algorithms, HISD has three characteristics that make it very attractive for practical systems. First, it performs very close to the ML limit. Second, it has a low computational complexity. Finally, The inherent parallel structure of the proposed algorithm allows us highly suitable hardware implementation. In this paper the HISD is efficiently employed in the case of MIMO (Mult...
ABSTRACT Notre application s'inscrit dans un contexte d'analyse de signaux radar.... more ABSTRACT Notre application s'inscrit dans un contexte d'analyse de signaux radar. Notre objectif est de concevoir un système permettant d'effectuer en temps réel une détection d'énergie en temps et en fréquence sur un signal large bande. Le CAN du système échantillonne le signal en fréquence intermédiaire à une fréquence de 1,33 GHz. La bande utile du signal est de 500 MHz. Les échantillons sont ensuites regroupés par paquet de 8 échantillons consécutifs et sont entrées dans le FPGA à une fréquence d'horloge de 167 MHz. Les échantillons sont codés sur 8 bits. Une première étape de démodulation numérique (non décrite dans ce résumé), permet de démoduler le signal et de générer les échantillons complexes en bande de base (4 échantillons par cycles d'horloge). L'étape suivante consiste à effectuer l'analyse temps fréquence en temps réel. Le principe est de diviser la bande du signal en N = 64 sous-bandes et de mesure l'énergie reçue dans chacune de ces sous-bandes. La réalisation d'une telle opération nécessite la création d'un banc de filtres.
La mobilite au niveau mondial necessite la gestion et la demodulation de nombreux standards tels ... more La mobilite au niveau mondial necessite la gestion et la demodulation de nombreux standards tels que GSM, IS95, DECT, le futur UMTS et les reseaux locaux RLAN. Il s'agit en effet de pouvoir communiquer en tout lieu, sur n'importe quel support et avec tout service voix, donnees de toute nature, diffusion. Les defis technologiques lies a cette RL sont de 2 ordres. Le premier concerne l'architecture. Le second defi concerne le logiciel. En effet rien ne sert de definir une architecture entierement reprogrammable si on ne peut pas l'alimenter avec les logiciels adequats.
The ML detection used in many communication applications reduces down to solving an integer least... more The ML detection used in many communication applications reduces down to solving an integer least-squares problem, which minimizes an objective function. If the matrix is not orthogonal, this problem is NP-hard. An exhaustive search over all feasible solutions is thus only applicable when the dimension of the problem is low. In the literature, there are two main available classes of methods. The first class comprises of exact methods whose complexities are roughly cubic in the dimension for medium to high SNR. However, the computational complexity remains exponential for low SNR. Moreover, these algorithms are iterative and thus, not well suited for hardware implementation. The second class is based on a relaxation strategy approach. However, the BER performance of this class is still not optimum. In this paper, we use a third class of methods for the resolution of this problem. Our approach, named Geometrical Diversification and Greedy Intensification (GDGI) associates two compleme...
Procedia Computer Science
2015 IEEE 12th International Conference on Wearable and Implantable Body Sensor Networks (BSN), 2015
Uploads
Papers by Christian Roland