Search

Proactive content caching has emerged as a promising solution to cope with exponentially increasing mobile data traffic. The popular user contents can be cached near the network edge for faster retrieval and processing. Current state-of-the-art approaches adopt a centralized model training mechanism that requires high communication and data exchange overheads to predict content popularity. Moreover, these approaches fail to deal with the dynamicity of the environment since they do not take into account the users’ mobility information and are unable to incorporate content offload timings. In this paper, we address these limitations by proposing a novel federated learning-based Mobility and Demand-aware Proactive Content Offloading (MDPCO) framework. MDPCO exploits distributed learning strategies and capitalizes on users’ mobility and demand information for proactive content offloading. Extensive simulations are carried out to validate the efficacy of MDCPO against local and cloud-based models. Our proposed model yields an average performance improvement of 6.7% in comparison to the cloud-based model. Furthermore, with the increase in the number of fog servers, the MDPCO achieves a 9.8% higher data offloading ratio and 1.18% increase in the downlink rates while being more energy-efficient than cloud-based approaches.

Non-speech emotion recognition has a wide range of applications including healthcare, crime control and rescue, and entertainment, to name a few. Providing these applications using edge computing has great potential, however, recent studies are focused on speech-emotion recognition using complex architectures. In this paper, a non-speech-based emotion recognition system is proposed, which can rely on edge computing to analyse emotions conveyed through non-speech expressions like screaming and crying. In particular, we explore knowledge distillation to design a computationally efficient system that can be deployed on edge devices with limited resources without degrading the performance significantly. We comprehensively evaluate our proposed framework using two publicly available datasets and highlight its effectiveness by comparing the results with the well-known MobileNet model. Our results demonstrate the feasibility and effectiveness of using edge computing for non-speech emotion detection, which can potentially improve applications that rely on emotion detection in communication networks. To the best of our knowledge, this is the first work on an edge-computing-based framework for detecting emotions in non-speech audio, offering promising directions for future research., Comment: Under review

Recent exponential growth of data over cellular networks has cause the progression from conventional mobile communication networks to heterogeneous cellular networks (HetNets). Quality of experience (QoE)-aware traffic load balancing in such dense HetNets is considered a major problem. Current HetNets exploit reactive load balancing schemes that hinder in achieving desired QoE gain. In this paper, we propose a novel proactive load balancing framework which leverages content caching and mobility prediction to improve user QoE. We use Semi-Markov model to predict users' future cells and a novel proactive caching algorithm is proposed to pre-fetch user future demands at these cells. This allows us to reduce cell congestion and offers better average downlink rates. To validate the effectiveness of the proposed framework, system level simulations are performed and compared with state of the art approaches. Encouragingly, our system achieves up-to $98.7\%$ cell load balancing in terms of Jain's fairness index, reduces $20\%$ backhaul load, and provides $31\%$ improved downlink rates at congested cells in contrast to reactive approaches., This is incomplete draft

Human lip-reading is a challenging task. It requires not only knowledge of underlying language but also visual clues to predict spoken words. Experts need certain level of experience and understanding of visual expressions learning to decode spoken words. Now-a-days, with the help of deep learning it is possible to translate lip sequences into meaningful words. The speech recognition in the noisy environments can be increased with the visual information [1]. To demonstrate this, in this project, we have tried to train two different deep-learning models for lip-reading: first one for video sequences using spatiotemporal convolution neural network, Bi-gated recurrent neural network and Connectionist Temporal Classification Loss, and second for audio that inputs the MFCC features to a layer of LSTM cells and output the sequence. We have also collected a small audio-visual dataset to train and test our model. Our target is to integrate our both models to improve the speech recognition in the noisy environment

Mobile phone based potable water quality assessment device is developed to analyze and study water pollution level at Indus river. Indus river is habitat of endangered Indus river dolphin and water pollution is one of major causes of survivability threats for this specie. We tested device performance at the six locations of Lahore canal. pH of canal water deviates from the normal range of the irrigation water. In future, we will study correlation between water pollution level and habitat usage of Indus river dolphin using water quality assessment device and hydrophone array based passive acoustic monitoring (PAM) system.

Fairness is the significant factor to sustain best effort delivery of network services. Now-a-days, real-time multimedia applications have evolved largely over the Internet. Most of multimedia services are unresponsive during network congestion. Unresponsive traffic streams steal most network bandwidth and starve out some other flows which are responsive during network congestion. In the presence of these unresponsive traffic flows, protection of responsive flows has become a major issue. Many Active Queue Management (AQM) based solutions have been recommended to protect responsive traffic flows from unresponsive ones and to ensure fairness among all traffic flows. The thesis proposes a novel AQM scheme CHOKeD, to deliver fairness among all flows of a congested link. It is a completely stateless approach. CHOKeD is based on dynamic drawing factor to penalize unresponsive traffic. It successfully protects responsive flows in the presence of unresponsive flows. CHOKeD features such as fairness, high throughput of responsive traffic and stateless design, are encouraging factors for its deployment over the edge as well as the network core routers. Extensive simulations have been carried out to evaluate its performance under real-time network scenarios, MS thesis