1: \begin{abstract}

2: Hybrid transceivers are designed for linear decentralized estimation

3: (LDE) in a mmWave multiple-input multiple-output (MIMO) IoT network

4: (IoTNe). For a noiseless fusion center (FC), it is demonstrated that

5: the MSE performance is determined by the number of RF chains used at

6: each IoT node (IoTNo). Next, the minimum-MSE RF transmit precoders

7: (TPCs) and receiver combiner (RC) matrices are designed for this setup

8: using the dominant array response vectors, and subsequently, a

9: closed-form expression is obtained for the baseband (BB) TPC at each IoTNo using

10: Cauchy's interlacing theorem. For a realistic noisy FC, it is shown

11: that the resultant mean squared error (MSE) minimization problem is

12: non-convex. To address this challenge, a block-coordinate

13: descent-based iterative scheme is proposed to obtain the fully digital

14: TPC and RC matrices followed by the simultaneous orthogonal matching

15: pursuit (SOMP) technique for decomposing the fully-digital transceiver

16: into its corresponding RF and BB components. A theoretical proof of

17: the convergence is also presented for the proposed iterative design

18: procedure. Furthermore, robust hybrid transceiver designs are also

19: derived for a practical scenario in the face of channel state

20: information (CSI) uncertainty. The centralized MMSE lower bound has

21: also been derived that benchmarks the performance of the proposed LDE

22: schemes. Finally, our numerical results characterize the performance

23: of the proposed transceivers as well as corroborate our various

24: analytical propositions.

25: \end{abstract}

26: