On-board satellite hyperspectral imagery (HSI) involves next-generation sensors with high data acquisition rates and limited availability of resources. This paper presents a band interleaved by line (BIL) architecture of a simple lossless algorithm (SLA) to extend the compatibility of SLA with the most widely used on-board interleaved HSI sensors. The architecture configures the data handling block of SLA based on the data dependencies and meets the neighbourhood requirements of the pre-processing block in the BIL order, which avoids any performance penalty arising from the reordering of incoming sensor data for a particular input sampling order. The region of operations for the pre-processing block of SLA are adapted to the BIL order and finally, the prediction residual is encoded using the fixed parameter Golomb-Rice (GR) entropy encoder. The architecture is implemented on the Kintex KCU-105 evaluation board using the Verilog hardware description language (HDL). The test datasets from the Consultative Committee for Space Data Systems (CCSDS) corpus of hyperspectral data are used to evaluate the architecture. The architecture achieves a maximum throughput of 4.848 Gbps for an input dynamic range of 16-bit, which shows significant improvement from the previous BSQ implementations of SLA and the BSQ and band interleaved implementations of the CCSDS 123.0-B-1 standard in terms of the resource utilization, power consumption, and the maximum frequency of operation.

On-board satellite remote sensing necessitates the data processing blocks to comprise high throughput with low computational complexity to match the high-speed data acquisition with the constraints of resource, power, timing, and downlink bandwidth. This paper presents a high-performance architecture of a simple lossless algorithm (SLA) for on-board satellite multispectral and hyperspectral data compression, which can be configured to all the directional modes of SLA in the band-sequential (BSQ) sampling order. The architecture supports a 16-bit input dynamic range and includes a mode synchronized data handling block along with pipelined implementations of the pre-processing and entropy encoding blocks. A novel adaptation is proposed in the entropy encoder block, which uses a correlation-based adaptation (CBA) in Golomb-Rice (GR) encoding to improve the compression performance of SLA while maintaining the low computational complexity. Kintex KCU-105 evaluation board is used for implementation of the architecture for testing with standard test datasets from the Consultative Committee for Space Data System (CCSDS) corpus of data for multispectral and hyperspectral imagery. The proposed architecture shows comparable compression performance to the low complexity mode of the CCSDS 123.0-B-1 standard and is within$\approx 1$bits per sample (bps) range in comparison to the default mode of the standard. A throughput of$\approx 4$Gbps is achieved at a clock frequency of 250 MHz with lesser resource utilization and power consumption compared to the CCSDS 123.0-B-1 standard implementations.

On-board satellite hyperspectral imagery (HSI) is characterized by a large amount of data acquisition with decreasing ground sample distance and limited resources on-board to meet relatively stable channel bandwidth requirements. This paper presents an efficient hardware architecture of a simple lossless algorithm (SLA) for on-board satellite hyperspectral data compression in band sequential (BSQ) order. The architecture uses an efficient data handling block along with the pipelined implementation of the pre-processing stage to improve the throughput. Kintex KCU-105 evaluation board is used to implement the architecture for testing with standard datasets from the Consultative Committee for Space Data System (CCSDS) corpus of data for HSI. Throughput, power, and resource utilization are compared with the hardware implementations of the CCSDS 123.0-B-1 standard. SLA shows real-time deployment capabilities by achieving a throughput of ~3.8 Gbps 238 MHz clock frequency with lesser resource utilization and affirm as a low complexity lossless alternative for on-board satellite hyperspectral data compression.

As resolution of on-board imaging spectrometer keeps improving, data acquisition rate increases and resource limited satellite environment necessitates for computationally simple data compression methods to meet timing, bandwidth and resource requirements with error resilience. This letter proposes a new lossless, prediction based algorithm for on-board satellite hyperspectral data compression that utilizes spectral as well as spatial correlation and at the same time, is computationally less complex. Concept of non-binary tree traversal is used with nearest neighbor method and implemented using neighbor driven decision making in pre-processing stage. Previously processed pixels are used to minimize the prediction residual, which makes more than 80% calculations causal in nature and thereby reducing the computational complexity of the algorithm. The prediction residual is then encoded using sample adaptive Golomb coding in band-sequential order. CCSDS corpus of data for hyperspectral images is used for evaluating the performance of the algorithm. The proposed method shows reduced computational complexity and lesser data dependencies compared to the CCSDS 123.0-B-1 standard when similar spectral vicinity is considered, and comparable compression performance compared to other state-of-the-art on-board lossless compression methods.

Sparse recovery algorithms are integral to compressed sensing (CS) as they facilitate the reconstruction of higher dimensional signals from sub-Nyquist measurements. Although Orthogonal Matching Pursuit (OMP) has been ubiquitously adopted in hardware implementations to curb the computational complexity of CS recovery, increasing sparsity levels incur higher reconstruction costs. Sparsity independent regularized pursuit (SIRP) is a recent algorithm that employs parallel index selection and regularization to curtail the number of iterations required to reconstruct the signal and thereby enhance the reconstruction speed. This paper proposes a novel reformulation of the SIRP algorithm from the hardware perspective, incorporating a cheaper regularization strategy and a modified Gram-Schmidt (MGS) based incremental QR decomposition (QRD) approach. The proposed design incorporates an iterative QRD architecture with feedback circuitry to exploit the parallelism of the triangularization step in MGS. Additionally, a fast inverse square block circumvents the need for parallel divider blocks giving considerable hardware and latency savings. The design reuses the iterative QRD block to implement the interdependent computations of the algorithm by sophisticated scheduling techniques. The proposed implementation on a Xilinx Virtex Ultrascale FPGA can recover 1024-dimensional signals from 25% measurements within$77~\mu \text{s}$, which translates to a 37% reduction in processing cycles from state-of-art.

Atmospheric motion vectors (AMVs) estimation helps in better understanding of atmospheric dynamics and also plays a key role in weather forecasting. It has been a challenging task because of the nonrigid motion of clouds and cyclones. In this paper, a modified Weighted Ensemble Transform Kalman Filter-based data assimilation technique is proposed for accurate flow vector estimation at each pixel directly from satellite generated infrared images of clouds/cyclones. This method provides clear visualization of both local and global motion with spatial and temporal consistencies very efficiently even in the case of splitting and merging of clouds or over long tracks. One of the key abilities of proposed method is in forecasting applications and also for generating motion vectors in the absence of data in real scenarios, even without the usage of the existing complex weather models. Estimated AMVs are validated using state-of-the-art European Centre for Medium-range Weather Forecasting (ECMWF) analysis data, and cyclone tracks are validated using the Indian Meteorological Department (IMD) best track data. The results obtained demonstrate the efficacy of proposed method over other existing methods.