Vinod Sharma

Prashant Narayanan, Vinod Sharma

We consider a scenario where N users try to access a common base station. Associated with each user is its channel state and a finite queue which varies with time. Each user chooses his power and the admission control variable in a dynamic manner so as to maximize his expected throughput. The throughput of each user is a function of the actions and states of all users. The scenario considers the situation where each user knows his channel and buffer state but is unaware of the states and actions taken by the other users. We consider the scenario when each user is saturated (i.e., always has a packet to transmit) as well as the case when each user is unsaturated. We formulate the problem as a Markov game and show connections with strategic form games. We then consider various throughput functions associated with the multiple user channel and provide algorithms for finding these equilibria.

Ramkumar Raghu, Mahadesh Panju, Vaneet Aggarwal et al.

Multicasting in wireless systems is a natural way to exploit the redundancy in user requests in a Content Centric Network. Power control and optimal scheduling can significantly improve the wireless multicast network's performance under fading. However, the model based approaches for power control and scheduling studied earlier are not scalable to large state space or changing system dynamics. In this paper, we use deep reinforcement learning where we use function approximation of the Q-function via a deep neural network to obtain a power control policy that matches the optimal policy for a small network. We show that power control policy can be learnt for reasonably large systems via this approach. Further we use multi-timescale stochastic optimization to maintain the average power constraint. We demonstrate that a slight modification of the learning algorithm allows tracking of time varying system statistics. Finally, we extend the multi-timescale approach to simultaneously learn the optimal queueing strategy along with power control. We demonstrate scalability, tracking and cross layer optimization capabilities of our algorithms via simulations. The proposed multi-timescale approach can be used in general large state space dynamical systems with multiple objectives and constraints, and may be of independent interest.

Ramkumar Raghu, Pratheek Upadhyaya, Mahadesh Panju et al.

We consider a multicast scheme recently proposed for a wireless downlink in [1]. It was shown earlier that power control can significantly improve its performance. However for this system, obtaining optimal power control is intractable because of a very large state space. Therefore in this paper we use deep reinforcement learning where we use function approximation of the Q-function via a deep neural network. We show that optimal power control can be learnt for reasonably large systems via this approach. The average power constraint is ensured via a Lagrange multiplier, which is also learnt. Finally, we demonstrate that a slight modification of the learning algorithm allows the optimal control to track the time varying system statistics.

V. Sriram Siddhardh Nadendla, Vinod Sharma, Pramod K. Varshney

In this paper, we model a complete-information zero-sum game between a centralized detection network with a multiple access channel (MAC) between the sensors and the fusion center (FC), and a jammer with multiple transmitting antennas. We choose error probability at the FC as the performance metric, and investigate pure strategy equilibria for this game, and show that the jammer has no impact on the FC's error probability by employing pure strategies at the Nash equilibrium. Furthermore, we also show that the jammer has an impact on the expected utility if it employs mixed strategies.

Shahid Mehraj Shah, Krishna Chaitanya A, Vinod Sharma

In this paper a $K$-user fading multiple access channel with and without security constraints is studied. First we consider a F-MAC without the security constraints. Under the assumption of individual CSI of users, we propose the problem of power allocation as a stochastic game when the receiver sends an ACK or a NACK depending on whether it was able to decode the message or not. We have used Multiplicative weight no-regret algorithm to obtain a Coarse Correlated Equilibrium (CCE). Then we consider the case when the users can decode ACK/NACK of each other. In this scenario we provide an algorithm to maximize the weighted sum-utility of all the users and obtain a Pareto optimal point. PP is socially optimal but may be unfair to individual users. Next we consider the case where the users can cooperate with each other so as to disagree with the policy which will be unfair to individual user. We then obtain a Nash bargaining solution, which in addition to being Pareto optimal, is also fair to each user. Next we study a $K$-user fading multiple access wiretap Channel with CSI of Eve available to the users. We use the previous algorithms to obtain a CCE, PP and a NBS. Next we consider the case where each user does not know the CSI of Eve but only its distribution. In that case we use secrecy outage as the criterion for the receiver to send an ACK or a NACK. Here also we use the previous algorithms to obtain a CCE, PP or a NBS. Finally we show that our algorithms can be extended to the case where a user can transmit at different rates. At the end we provide a few examples to compute different solutions and compare them under different CSI scenarios.

Sahil Sharma, Vinod Sharma, Atul Sharma

Areas where Artificial Intelligence (AI) & related fields are finding their applications are increasing day by day, moving from core areas of computer science they are finding their applications in various other domains.In recent times Machine Learning i.e. a sub-domain of AI has been widely used in order to assist medical experts and doctors in the prediction, diagnosis and prognosis of various diseases and other medical disorders. In this manuscript the authors applied various machine learning algorithms to a problem in the domain of medical diagnosis and analyzed their efficiency in predicting the results. The problem selected for the study is the diagnosis of the Chronic Kidney Disease.The dataset used for the study consists of 400 instances and 24 attributes. The authors evaluated 12 classification techniques by applying them to the Chronic Kidney Disease data. In order to calculate efficiency, results of the prediction by candidate methods were compared with the actual medical results of the subject.The various metrics used for performance evaluation are predictive accuracy, precision, sensitivity and specificity. The results indicate that decision-tree performed best with nearly the accuracy of 98.6%, sensitivity of 0.9720, precision of 1 and specificity of 1.

Shahid M. Shah, Vinod Sharma

Security constraint results in \textit{rate-loss} in wiretap channels. In this paper we propose a coding scheme for two user Multiple Access Channel with Wiretap (MAC-WT), where previous messages are used as a key to enhance the secrecy rates of both the users until we achieve the usual capacity region of a Multiple Access Channel (MAC) without the wiretapper (Shannon capacity region). With this scheme all the messages transmitted in the recent past are secure with respect to all the information of the eavesdropper till now. To achieve this goal we introduce secret key buffers at both the users, as well as at the legitimate receiver (Bob). Finally we consider a fading MAC-WT and show that with this coding/decoding scheme we can achieve the capacity region of a fading MAC (in ergodic sense).

Shahid M. Shah, Vinod Sharma

Reliable communication imposes an upper limit on the achievable rate, namely the Shannon capacity. Wyner's wiretap coding, which ensures a security constraint also, in addition to reliability, results in decrease of the achievable rate. To mitigate the loss in the secrecy rate, we propose a coding scheme where we use sufficiently old messages as key and for this scheme prove that multiple messages are secure with respect to (w.r.t.) all the information possessed by the eavesdropper. We also show that we can achieve security in the strong sense. Next we consider a fading wiretap channel with full channel state information of the eavesdropper's channel and use our coding/decoding scheme to achieve secrecy capacity close to the Shannon capacity of the main channel (in the ergodic sense). Finally we also consider the case where the transmitter does not have the instantaneous information of the channel state of the eavesdropper, but only its distribution.

Shahid Mehraj Shah, Vinod Sharma

We consider a two user multiple-access channel with an eavesdropper at the receiving end. We use previously transmitted messages as a key in the next slot till we achieve the capacity region of the usual Multiple Access Channel (MAC).

Shahid Mehraj Shah, Parameswaran S, Vinod Sharma

We consider a wiretap channel and use previously transmitted messages to generate a secret key which increases the secrecy capacity. This can be bootstrapped to increase the secrecy capacity to the Shannon capacity without using any feedback or extra channel while retaining the strong secrecy of the wiretap channel.

Shahid Mehraj Shah, Vinod Sharma

In this paper we consider a wiretap channel with a secret key buffer. We use the coding scheme of [1] to enhance the secrecy rate to the capacity of the main channel, while storing each securely transmitted message in the secret key buffer. We use the oldest secret bits from the buffer to be used as a secret key to transmit a message in a slot and then remove those bits. With this scheme we are able to prove stronger results than those in [1]. i.e., not only the message which is being transmitted currently, but all the messages transmitted in last $N_1$ slots are secure with respect to all the information that the eavesdropper possesses, where $N_1$ can be chosen arbitrarily large.