Wassim Jouini

Multi-armed bandit based policies for cognitive radio's decision making issues

We suggest in this paper that many problems related to Cognitive Radios (CR) decision making inside CR equipments can be formalized as Multi-Armed Bandit problems and that solving such problems by using Upper Confidence Bound (UCB) algorithms can lead to high-performance CR devices. An application of these algorithms to an academic Cognitive Radio problem is reported.

Upper Confidence Bound Based Decision Making Strategies and Dynamic Spectrum Access

In this paper, we consider the problem of exploiting spectrum resources for a secondary user (SU) of a wireless communication network. We suggest that Upper Confidence Bound (UCB) algorithms could be useful to design decision making strategies for SUs to exploit intelligently the spectrum resources based on their past observations. The algorithms use an index that provides an optimistic estimation of the availability of the resources to the SU. The suggestion is supported by some experimental results carried out on a specific dynamic spectrum access (DSA) framework.

Energy Detection Limits Under Log-Normal Approximated Noise Uncertainty

We revisit, in this letter, the impact of noise uncertainty on the performance of the well known energy detector. Mainly, we reconsider the case of a Log-Normal approximated noise uncertainty suggested in the work of Alexander Sonnenschein and Philip M. Fishman. We show that under a Log-Normal noise uncertainty, closed form expressions of the detectors performances and limits can be provided. Thus we show that, relying on mild approximations, we can design a detector with a fixed probability of false alarm function of the uncertainty, and present a new expression of the SNR-wall that depends on the desired performances of the detector as well as the introduced uncertainty parameter.

Improving Case-Based Reasoning Systems by Combining K-Nearest Neighbour Algorithm with Logistic Regression in the Prediction of Patients’ Registration on the Renal Transplant Waiting List

Introduction Case-based reasoning (CBR) is an emerging decision making paradigm in medical research where new cases are solved relying on previously solved similar cases. Usually, a database of solved cases is provided, and every case is described through a set of attributes (inputs) and a label (output). Extracting useful information from this database can help the CBR system providing more reliable results on the yet to be solved cases. Objective We suggest a general framework where a CBR system, viz. K-Nearest Neighbour (K-NN) algorithm, is combined with various information obtained from a Logistic Regression (LR) model, in order to improve prediction of access to the transplant waiting list. Methods LR is applied, on the case database, to assign weights to the attributes as well as the solved cases. Thus, five possible decision making systems based on K-NN and/or LR were identified: a standalone K-NN, a standalone LR and three soft K-NN algorithms that rely on the weights based on the results of the LR. The evaluation was performed under two conditions, either using predictive factors known to be related to registration, or using a combination of factors related and not related to registration. Results and Conclusion The results show that our suggested approach, where the K-NN algorithm relies on both weighted attributes and cases, can efficiently deal with non relevant attributes, whereas the four other approaches suffer from this kind of noisy setups. The robustness of this approach suggests interesting perspectives for medical problem solving tools using CBR methodology.

Blind standard identification with bandwidth shape and GI recognition using USRP platforms and SDR4all tools

In this paper, focusing on identifying standards blindly, we propose a bandwidth shape sensor and a GI (guard interval) sensor using USRP (Universal Software Radio Peripheral) platforms and SDR4all tools. These sensors are fundamental parts of the so-called Blind Standard Recognition Sensor. The blind standard bandwidth sensor is based on a Radial Basis Function Neuronal Network designed in Matlab. We have presented first experience of using blind standard bandwidth sensor in a previous work. We will provide in this paper further details on the results of this sensor (simulations, preliminary implementations and validations). The GI sensor is implemented in order to improve the detection performance in the case of two identical bandwidth shapes. The SDR4all driver offers a simple yet efficient interface between the Matlab signal processing codes and the USRP transmitting and receiving platforms. These simple and easily accessible software defined radio tools were used to design and implement two sensors. The inducted simulations and experiments show that the designed system is indeed able to discriminate three standard-like spectrums (e.g., GSM-like, UMTS-like and OFDM-like) under simple yet real transmission conditions using their different bandwidth shapes and to identify a GI-OFDM-like system using cyclic autocorrelation method.

Log-normal approximation of chi-square distributions for signal processing

We investigate in this paper a Log-Normal approximation of X2 distributions. Our study is motivated by the analysis of ratios of random variables where both, Log-Normal and X2 distributions are involved. Such ratios appear, for instance, when dealing with an energy detector while only an imperfect knowledge of the noise level is available. In order to characterize the distribution of the ratio, an accurate approximation of the X2 distribution by a Log-Normal distribution would highly simplify this problem known to be analytically intractable otherwise.

Blind Bandwidth Shape Recognition for Standard Identification Using USRP Platforms and SDR4all Tools

In this paper a blind standard bandwidth shape sensor is implemented on a USRP (Universal Software Radio Peripheral) platform using SDR4all tools. This sensor is a fundamental part of the so-called Blind Standard Recognition Sensor. The blind standard bandwidth sensor is based on a Radial Basis Function Neuronal Network designed in Matlab. The SDR4all driver offers a simple yet efficient interface between the Matlab signal processing codes and the USRP transmitting and receiving platforms. To the best of our knowledge, it is the first time that simple and easily accessible software defined radio tools were used to design and implement this sensor. Although the experiments were realized under line-of-sight transmission conditions the results show that the designed system is indeed able to discriminate several standard-like spectrums under real transmission conditions using their different bandwidth shapes.

Cognitive radio equipments supporting spectrum agility

This paper describes the practical design issues and potential strategies for implementing cognitive radio (CR) equipments that support spectrum agility. We first discuss a simplified cognitive cycle that satisfies the requirements of a typical CR scenario. We give a brief overview of the various steps involved in the above cycle, namely, sensing, decision making and flexible signal processing. A management architecture called HDCRAM (Hierarchical and Distributed Cognitive Radio Architecture Management) is proposed as a solution for efficiently managing all the steps in the cognitive cycle. Finally we give an integrated proposal, which serves a starting point for designing future spectrum agile CR equipments.