The evolution towards ubiquitous (anytime, anywhere) communications and computing poses problems requiring novel ways of utilizing the frequency spectrum and the wireless channel. The group’s focus is on these aspects
Modelling of optical carrier recovery and phase synchronization scheme
Modelling of optical carrier recovery and phase synchronization scheme. Current optical fiber communication systems are deviating from intensity modulated transmission schemes to phase modulated transmission schemes due to various reasons such as impairment mitigation, data rate increment, etc. Therefore, this research focuses on extracting phase information from a degraded phase modulated signal, which will be useful in optical detection and regeneration schemes. This research is funded by National Research Council (NRC) research grants.
Polarization Insensitive, Phase sensitive amplifier for phase Regeneration
New standards have been released recently for increasing the data rates used in optical fiber core and metro networks operating at 100 Gbps. Basically 100 Gbps systems employ DP-QPSK schemes and in some of the 40 Gbps optical networks are operating in (D)PSK/QPSK schemes. Therefore, this research focuses on extracting phase information from a degraded phase modulated signal which will be useful in optical detection and regeneration schemes. This research is funded by National Research Council (NRC) research grants.
The low penetration of on-board devices supporting Vehicle-to-vehicle (V2V) communications hinders many possible applications in intelligent transportation systems. The research focuses on using communication capabilities of mobile phones to facilitate the process, and design low cost on-board units with much of the V2V communications processing handed over to the mobile phone. This research is funded by Senate Research Committee (SRC) long and medium term grants.
The detection of signals in noisy observations
The detection of signals in noisy observations is one of the fundamental problems in statistical signal processing. This problem also arises in various other scientific disciplines such as radar, sonar, wireless communications and finance. In its most basic form, the presence of a signal amounts to rank one departure of the population covariance matrix from the identity. Equivalently, the largest eigenvalue of the population covariance matrix deviates from unity. Since we do not have access to the population covariance matrix,we focus on the largest eigenvalue of the sample covariance matrix (i.e., signal plus noise) formed with the noisy observations (say S). Moreover, if the noise co-variance matrix is unknown, then it is common to construct another sample covariance matrix from noise only observations (sayR). Then it is natural to consider the behavior of the largest eigenvalue of F = R-1S in order to infer the presence of a signal. Therefore, the main objective of this project is to investigate the asymptotic (i.e., high dimensional) behavior of the largest eigenvalue of F matrix when R and S are Wishart distributed.