Loading...

Change Point Detection in Molecular Carrier Based Nano Networks

Ghoroghchian, Nafiseh | 2017

2640 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49801 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Nasiri Kenari, Masoumeh; Aminzadeh Gohari, Amin
  7. Abstract:
  8. Molecular communication (MC) is an emerging communication paradigm, whereas molecules are used as information carriers to establish communication among elements in nano-meter to meter scales. In this thesis, we investigate the problem of detecting and monitoring changes (abnormality) based on molecular communication, using quickest change point detection scheme. We assume the distributions and parameters of the system are known. To this end, we consider a network of multiple sensors, each sensing its surrounding and employing On-Off-keying modulation for data transmission toward a fusion center (FC). An abnormality initiates randomly in time and location, and further propagates in the environment in an stochastic manner. The objective is to watch and make decisions on the time and location of changes in the environment based on the received observations from sensors. Such assumptions call for considering spatial and temporal correlations among different and same sensors’ transmitting signals. In this thesis, unlike wireless sensor networks, in which correlations are modeled with normal distribution using different forms of covariance matrix, the framework of Partially Observable Markov Decision Processes (POMDPs) based on non-homogeneous Markov models, is proposed. This framework enables system to include any correlation form in order to better model practical systems and applications. Optimum metrics for both change detection (stopping-time) and monitoring scenarios are presented and the corresponding detectors are obtained. The metric in stopping-time (detection) scenario is to minimize the delay of announcing an abnormality occurrence subject to constraints on false alarm and missed identification probabilities. Since the optimum detector is very complicated, by utilizing myopic policy, sub-optimum detectors with less complexity and acceptable performance are proposed. For monitoring scenario, the stopping-time problem is modified with the goal to determine how the changes are spread in time. Accordingly, one optimum and two sub-optimum metrics are defined and their corresponding detectors are derived. In this scenario, the detectors make the decision only in the time slots whereas a reliable decision can be made. The defined metrics minimize waiting-time, the time that the detector does not make decision due to lack of assurance, subject to constraints on false alarm and missed identification probabilities. For performance evaluation, we consider a previously reported system of tumor growth in a tissue. Using the general framework introduced in this thesis, we mathematically model the system and evaluate the performance of the proposed detectors. The examples considered demonstrate that the proposed detectors perform better compared with the non-parametric and also previously introduced detectors. Keywords: Multi-sensor data fusion network, Qui
  9. Keywords:
  10. Partially Observable Markov Decision Process ; Molecular Communication ; Monitoring ; Non Homogenious Markov Model ; MYOPIC Algorithm ; Multi-Sensor Data Fusion Network ; Quickest Change Detection ; Stopping Time

 Digital Object List

 Bookmark

...see more