The Modares Journal of Electrical Engineering
Volume:13 Issue: 2, 2013

The fusion of valuable spectral and spatial features can significantly improve the performance of high resolution hyperspectral images classification. In this paper, we propose a spectral and spatial feature extraction method based on discriminant analysis. To increase the class discrimination, we maximize the between-class scatters and minimize the within-class scatters. To include the spatial information in the feature extraction process, we estimate the spatial scatters in a spatial neighborhood window with multi-scale fashion. We compare our proposed method, which is called spectral-spatial discriminant analysis (SSDA), with some spatial-spectral feature extraction methods included original spectral bands plus Gabor filters, gray level co-occurance matrix (GLCM), and morphology profiles and also with some popular spectral feature extraction methods such as nonparametric weighted feature extraction (NWFE) and locality preserving projection (LPP). Moreover, we compare SSDA with some recently proposed spectral-spatial classification approaches. The experimental results on two real hyperspectral images show the good performance of SSDA compared to the competitor methods.

The problem of power allocation for distributed detection in a power constrained wireless sensor network is considered. The sensors are assumed to have independent observations and send their local decisions to a fusion center over multiple access channels. The Jensen-Shannon divergence between the distributions of the detection statistic under different hypotheses is used as a performance criterion. It is demonstrated that by applying the proposed measure power allocation is more efficient comparing to other criteria like mean square error or Jeffry divergence.

In this paper, we formulate the resource allocation problem for secure communications in a single cell OFDMA network where a base station wants to communicate to the users which are spread over its coverage area while an adversary is present in the network. In our model, we assume that the adversary s an active attacker which is able to either eavesdrop on the information transmitted over the wireless link or jam the connection by sending an interference signal but is not able to both eavesdrop and jam at the same time. The attacker's aim is to choose between eavesdropping and jamming a subcarrier such that the secrecy rate over that subcarrier is minimized. Knowing this strategy, users should consider the minimum of secrecy data rate in both cases as their achievable secrecy rate over each subcarrier. We formulate our proposed resource allocation problem into an optimization problem whose objective is to maximize the total secrecy rate of all users over all subcarriers while it is required that the total transmit power of base station remains below a predefined threshold. To solve the proposed optimization problem, we use the dual Lagrange approach. Using simulations, we study the behavior and the efficiency of the proposed scheme.

One of the main challenges in tracking radio occultation (RO) signals of open loop approach is the excess Doppler estimation accuracy in lower troposphere. Propagation of RO signals through the lower troposphere with severe refractivity gradient results in high phase acceleration and low signal to noise ratio (SNR) signal. Because of the high refractivity in the lower moist troposphere, the received RO signals by global navigation satellite system (GNSS) receivers can have large excess Doppler which may vary rapidly. Due to limitation on bandwidth and transmitted information to post processing, we cant use very high sampling rate at GNSS receivers at LEO satellite. Therefore, try to enhance the accuracy of the Doppler predicted models in the satellite makes the GNSS-RO receivers by using efficiently frequency estimation method at GNSS receivers. In this regard, we investigate the event of RO signal and then the various frequency estimation methods to improve the estimates of the Doppler frequency in post processing are evaluated. The various frequency estimation methods are investigated from performance and computational complexity perspective. Via simulation, the excess Doppler estimation accuracy in post processing for different frequency estimation methods are compared by its root mean squared error. Based on simulation results, it is seen that both ESPRIT and Jacobsen with Bias schemes have the better performance and the estimation error of them are less than that of other methods.

Elastic optical network has recently raised immense research efforts to serve as a highly-flexible and scalable platform required by the emerging on-demand cloud applications. Besides, virtual network embedding on top of elastic optical network has been introduced as a promising solution to fine-grain sharing of node processing and optical bandwidth resources. Owing to the diverse delay sensitivity requirements of heterogeneous cloud applications, an efficient double queue scheduler is presented, to deliver differentiated setup delay embedding in a fair manner. Extensive numerical simulations carried out over two prototype substrate networks, demonstrated that our proposed scheme can concomitantly enhance blocking and fairness performance of the single queue scheduler over wide range of offered traffic loads.

Nowadays, the spectrum is becoming more and more scarce because of not only the demand increase for wireless services and applications but also inflexible and rigid spectrum allocation policies. A promising solution to overcome this challenge is to exploit spectrum holes for secondary users in cognitive radio. Taking advantage of cognitive radio and dynamic spectrum access improves the efficiency of spectrum utilization as it allows secondary users to exploit the available frequency bandwidth. In an opportunistic spectrum access scenario, the secondary users can only utilize the frequency bandwidth when the primary user is inactive. In this paper, we assume that secondary users exploit the available frequency bandwidth using opportunistic spectrum access approach. As a result, one may model the available bandwidth to the secondary users as a random process taking value of 0 or 1. In this paper, we investigate the effects of Rayleigh fading environment on the capacity of AWGN point-to-point channel. Afterwards, we also go beyond point to point channels and analyze Multiple Access Channels. More precisely, we derive the closed form expression of ergodic capacity region of Multiple Access fading Channel with two independent sources. To the best of our knowledge, no prior work has investigated the ergodic capacity region of Multiple Access Channel under Rayleigh fading with the aforementioned approach. Numerical analysis are conducted to validate our theoretical results. Interestingly, numerical results indicate that, against the common sense, capacity of Rayleigh fading environment channels are higher than deterministic ones in Gaussian point-to-point channels.

In a framework where synchronization issues are taken into account, we examine the performance of a cyclostationary detection scheme proposed for interleaved single carrier frequency division multiple access (SCFDMA). The detection method under investigation here exploits the quasi-periodicity property of interleaved SCFDMA signal. Inserting synchronization pilot signals in the data stream might deteriorate this quasi-periodicity property. Although the detection method is afflicted by pilot signals and loses performance, we show that it still outperforms energy detection and cyclic prefix based detection. We also investigate the performance of the detector when the system is hit by frequency offset. We derive the distribution of the detection metric in the presence of frequency offset. When frequency offset is small enough, we prove that it does not change the metric distribution. We also show that for arbitrary practical frequency offsets, the performance degradation is still negligible.