Research

Federated Learning for Wireless Communication Systems

:

  1. Communication-Efficient Federated Learning for Massive MIMO

    Publications:

    1. S. Jha, AKR Chavva, S. Kadambar, "Communication-efficient Federated Learning for CSI Feedback in Massive MIMO," submitted to 2023 IEEE WCNC .

Distributed Optimization

:

  1. Over-the-Air Distributed Optimization (joint work with Prathmesh Mayekar , Dr. Himanshu Tyagi )

    Publications:

    1. S. Jha, "Fundamental Limits of Distributed Optimization over Multiple Access Channels," under review at IEEE Transactions on Communications. [link]
    2. S. Jha, P. Mayekar, "Fundamental Limits of Distributed Optimization over Multiple Access Channels," in proceedings 2023 IEEE Information Theory Workshops (ITW), Saint-Malo, France, 2023, pp. 406-411, doi: 10.1109/ITW55543.2023.10161622. [paper]
    3. S. Jha, P. Mayekar, H. Tyagi, "Fundamental limits of over-the-air optimization: Are analog schemes optimal?," in proceedings 2021 IEEE Global Conference on Communications (GLOBECOM), Madrid, Spain. [paper]
    4. S. Jha, P. Mayekar, H. Tyagi, "Fundamental limits of over-the-air optimization: Are analog schemes optimal?," in IEEE Journal on Selected Areas of Information Theory, vol. 3, no. 2, pp. 217-228, June 2022. [paper]
  2. Information-Constrained Smooth Distributed Optimization

    Publications:

    1. P. Mayekar, S. Jha, A.T. Suresh, H. Tyagi, "Wyner-Ziv Estimators for Distributed Mean Estimation with Side Information and Optimization," in IEEE Transactions on Information Theory, Sept. 2023, doi: 10.1109/TIT.2023.3315719. [paper]
    2. P. Mayekar, S. Jha, H. Tyagi, "Wyner-Ziv compression is (almost) optimal for distributed optimization," in proceedings 2022 IEEE Symposium on Information Theory (ISIT), Helsinki, Finland. [paper]

Distributed Quantization

  1. Practical Wyner-Ziv Codes for Universal Quantization [Jan'19-Mar'22] (joint work with Dr. Himanshu Tyagi)

    :

    Publications:

    1. S. Jha, H. Tyagi, "Universal interactive Gaussian quantization with side information," in proceedings 2020 IEEE Information Theory Workshop (ITW), Italy. [paper]
    2. S. Jha, "Universal Gaussian quantization with side information using Polar lattices," in IEEE Journal on Selected Areas of Information Theory. [paper]

Other Projects

  1. Distributed Markov Chain Monte-Carlo under Communication Constraints [Jul'20-Oct'21] (joint work with Dr. Osvaldo Simeone (KCL, UK) , Dr. Himanshu Tyagi)

    :

  2. Design of Polar decoders for short-to-medium blocklengths [Jan'18-Apr'19] (joint work with Dr. Kuntal Deka (IIT Guwahati) , Dr. Shilpa (IIITG))

    :

    Publications:

    1. S. Jha, K. Deka, S. Rao, “Mutual-Information-Based Successive Cancellation List Decoding of Polar Codes”, 2019 IEEE 89th Vehicular Technology Conference: VTC2019-Spring, Kuala Lumpur, Malaysia. [paper]