Recent Submissions

  • Deep Learning Based Frequency-Selective Channel Estimation for Hybrid mmWave MIMO Systems

    Abdallah, Asmaa (2022-03-22) [Poster]
    Millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) systems typically employ hybrid mixed signal processing to avoid expensive hardware and high training overheads. {However, the lack of fully digital beamforming at mmWave bands imposes additional challenges in channel estimation. Prior art on hybrid architectures has mainly focused on greedy optimization algorithms to estimate frequency-flat narrowband mmWave channels, despite the fact that in practice, the large bandwidth associated with mmWave channels results in frequency-selective channels. In this paper, we consider a frequency-selective wideband mmWave system and propose two deep learning (DL) compressive sensing (CS) based algorithms for channel estimation.} The proposed algorithms learn critical apriori information from training data to provide highly accurate channel estimates with low training overhead. In the first approach, a DL-CS based algorithm simultaneously estimates the channel supports in the frequency domain, which are then used for channel reconstruction. The second approach exploits the estimated supports to apply a low-complexity multi-resolution fine-tuning method to further enhance the estimation performance. Simulation results demonstrate that the proposed DL-based schemes significantly outperform conventional orthogonal matching pursuit (OMP) techniques in terms of the normalized mean-squared error (NMSE), computational complexity, and spectral efficiency, particularly in the low signal-to-noise ratio regime.
  • Antenna-on-Chip with Ultra-Thin AMC for mm-Wave System-on-Chip Application

    Yu, Yiyang (2022-03-22) [Poster]
    Millimeter-wave System-on-Chip (SoC) has become an attractive approach to achieve highly-integrated wideband wireless systems. However, the Antenna-on-Chip (AoC) suffers from poor radiation due to the lossy silicon substrate in standard CMOS processes. An Artificial Magnetic Conductor (AMC), completely realized within the thin oxide layer (~10-15 ?m) and with the ground plane above the silicon substrate, can enhance the overall antenna gain by isolating the lossy silicon. However, fitting the AMC layers in this thin oxide, along with the AoC, is extremely challenging for the mm-wave spectrum, particularly for frequencies below 100 GHz. In this work, the thickness of the AMC was reduced by Embedded Guiding Structures (EGS), which utilizes the available metal layers, typically available in an AoC stack-up. An in-house CMOS compatible fabrication process has been used to realize the AoC, where typical low-conductivity adhesion layers have been avoided. In their replacement, surface roughness has been used in a unique way to provide the required adhesion between the layers. It is found that with EGS, the thickness of AMC can be reduced 41%. The AMC with EGS approach fits within an oxide of thickness 16 ?m. A monopole antenna, on top of this AMC, demonstrates a bandwidth of 6 GHz, a gain of 5.85 dBi and radiation efficiency of 57% at 94 GHz.
  • Graphene Based Geometric Diode for THz Communication

    Wang, Heng (2022-03-22) [Poster]
    For communication system, diodes are commonly used as clippers, gates, data modulators, etc. When the frequency goes up to THz band, a diode with an ultra-fast speed is essential. While traditional semiconductor-based diodes cannot work at such a high frequency, a novel diode named geometric diode can be a perfect candidate.
  • Scalable GPU-based Decoding Approach for Massive MIMO Technology

    Dabah, Adel (2022-03-22) [Poster]
    Large Multiple-Input Multiple-Output (L-MIMO) technology incorporates hundreds of antennas in transceivers to increase the data rate and reliability of the communication network. Enabling this technology requires a detection algorithm that ensures scalability in terms of the number of antennas while maintaining a real-time requirement and good error rate performance. To this aim, we propose in this paper a GPU-based multi-level detection approach for large MIMO systems. Our approach is based on a search tree representing all possible combinations of the transmitted signal. To target the optimal path in the search tree, our approach combines multiple tree levels and then chooses the best combination to extend one path P. The algorithm repeats the same process for the following levels until reaching a complete path. Therefore, it constructs only one complete path. The more level we combine, the more accurate our approach is. However, it also increases the complexity. To maintain a practical time complexity, we cast the computation at each iteration as matrix multiplication and then rely on the compute power of Graphics processing unit (GPU) accelerators for latency requirements. The obtained results show the scalability potential of our approach by dealing with up to 400 by 400 antennas and 64 QAM modulation under the practical time complexity of 10 ms while maintaining a good symbol error rate performance 10?2 at 23 dB. To our knowledge, such a balance between scalability, error rate, and complexity has never been achieved in the literature. Additionally, our GPU-based multi-level approach shows a 200× relative speedup compared to a similar serial CPU implementation.
  • LOCATION AWARENESS IN HIGH SPEED HYBRID-INTERNET OF UNDERWATER THINGS

    Saeed, Nasir (2022-03-22) [Poster]
    Localization of sensor nodes in the High speed internet of underwater things (IoUT) is of considerable significance due to its various applications, such as navigation, data tagging, and detection of underwater objects. Therefore, in this paper, we propose a hybrid Bayesian multidimensional scaling (BMDS) based localization technique that can work on a fully hybrid IoUT network where the nodes can communicate using either optical, magnetic induction, and acoustic technologies. These communication technologies are already used for communication in the underwater environment; however, lacking localization solutions. Optical and magnetic induction communication achieves higher data rates for short communication. On the contrary, acoustic waves provide a low data rate for long-range underwater communication. The proposed method collectively uses optical, magnetic induction, and acoustic communication-based ranging to estimate the underwater sensor nodes' final locations. Moreover, we also analyze the proposed scheme by deriving the hybrid Cramer-Rao lower bound (H-CRLB). Simulation results provide a complete comparative analysis of the proposed method with the literature.