Fourth-Order Hierarchical Array: A Novel Scheme for Sparse Array Design Based on Fourth-Order Difference Co-Array

Conventional array designs based on circular fourth-order cumulant typically adopt a single expression form of the fourth-order difference co-array (FODCA), which limits the achievable degrees of freedom (DOFs) and neglects the impact of mutual coupling among physical sensors. To address above issues, this paper proposes a novel scheme to design arrays with increased DOFs by combining different forms of FODCA while accounting for mutual coupling. A novel fourth-order hierarchical array (FOHA) based on different forms of FODCA is constructed using an arbitrary generator set. The analytical expression between the coupling leakage of the generator and the resulting FOHA is derived. Two specific FOHA configurations are presented with closed-form sensor placements. The arrays not only offer increased DOFs for resolving more sources in direction of-arrival (DOA) estimation but also effectively suppress mutual coupling. Moreover, the redundancy of FODCA is examined, and it is shown that arrays based on the proposed scheme achieve lower redundancy compared to existing arrays based on FODCA. Meanwhile, the necessary and sufficient conditions for signal reconstruction by FOHA are derived. Compared with existing arrays based on FODCA, the proposed arrays provide enhanced DOFs and improved robustness against mutual coupling. Numerical simulations verify that FOHAs achieve superior DOA estimation performance compared with other sparse linear arrays.

A Novel Sparse Sum and Difference Co-Array With Low Redundancy and Enhanced DOF for Non-Circular Signals

Array structures based on the sum and difference co-arrays provide more degrees of freedom (DOF). However, since the growth of DOF is limited by a single case of sum and difference co-arrays, the paper aims to design a sparse linear array (SLA) with higher DOF via exploring different cases of second-order cumulants. We present a mathematical framework based on second-order cumulant to devise a second-order extended co-array (SO-ECA) and define the redundancy of SO-ECA. Based on SO-ECA, a novel array is proposed, namely low redundancy sum and difference array (LR-SDA), which can provide closed-form expressions for the sensor positions and enhance DOF in order to resolve more signal sources in the direction of arrival (DOA) estimation of non-circular (NC) signals. For LR-SDA, the maximum DOF under the given number of total physical sensors can be derived and the SO-ECA of LR-SDA is hole-free. Further, the corresponding necessary and sufficient conditions of signal reconstruction for LR-SDA are derived. Additionally, the redundancy and weight function of LR-SDA are defined, and the lower band of the redundancy for LR-SDA is derived. The proposed LR-SDA achieves higher DOF and lower redundancy than those of existing DCAs designed based on sum and difference co-arrays. Numerical simulations are conducted to verify the superiority of LR-SDA on DOA estimation performance and enhanced DOF over other existing DCAs.

FOGNA: An effective Sum-Difference Co-Array Design Based on Fourth-Order Cumulants

Array structures based on the fourth-order difference co-array (FODCA) provide more degrees of freedom (DOF). However, since the growth of DOF is limited by a single case of fourth-order cumulant in FODCA, this paper aims to design a sparse linear array (SLA) with higher DOF via exploring different cases of fourth-order cumulants. This paper presents a mathematical framework based on fourth-order cumulant to devise a fourth-order extend co-array (FOECA), which is equivalent to FODCA. A novel SLA, namely fourth-order generalized nested array (FOGNA), is proposed based on FOECA to provide closed-form expressions for the sensor locations and enhance DOF to resolve more signal sources in direction of arrival (DOA) estimation. FOGNA is consisted of three subarrays, where the first is a concatenated nested array and the other two subarrays are SLA with big inter-spacing between sensors. When the total physical sensors of FOGNA are given, the number of sensors in each subarray is determined by the designed method, which can obtain the maximum DOF under the proposed array structure and derive closed-form expressions for the sensor locations of FOGNA. The proposed array structure not only achieves higher DOF than those of existing FODCAs but also reduces mutual coupling effects. Numerical simulations are conducted to verify the superiority of FOGNA on DOA estimation performance and enhanced DOF over other existing FODCAs.

BadHMP: Backdoor Attack against Human Motion Prediction

Precise future human motion prediction over subsecond horizons from past observations is crucial for various safety-critical applications. To date, only one study has examined the vulnerability of human motion prediction to evasion attacks. In this paper, we propose BadHMP, the first backdoor attack that targets specifically human motion prediction. Our approach involves generating poisoned training samples by embedding a localized backdoor trigger in one arm of the skeleton, causing selected joints to remain relatively still or follow predefined motion in historical time steps. Subsequently, the future sequences are globally modified to the target sequences, and the entire training dataset is traversed to select the most suitable samples for poisoning. Our carefully designed backdoor triggers and targets guarantee the smoothness and naturalness of the poisoned samples, making them stealthy enough to evade detection by the model trainer while keeping the poisoned model unobtrusive in terms of prediction fidelity to untainted sequences. The target sequences can be successfully activated by the designed input sequences even with a low poisoned sample injection ratio. Experimental results on two datasets (Human3.6M and CMU-Mocap) and two network architectures (LTD and HRI) demonstrate the high-fidelity, effectiveness, and stealthiness of BadHMP. Robustness of our attack against fine-tuning defense is also verified.

A Hierarchical Dataflow-Driven Heterogeneous Architecture for Wireless Baseband Processing

Wireless baseband processing (WBP) is a key element of wireless communications, with a series of signal processing modules to improve data throughput and counter channel fading. Conventional hardware solutions, such as digital signal processors (DSPs) and more recently, graphic processing units (GPUs), provide various degrees of parallelism, yet they both fail to take into account the cyclical and consecutive character of WBP. Furthermore, the large amount of data in WBPs cannot be processed quickly in symmetric multiprocessors (SMPs) due to the unpredictability of memory latency. To address this issue, we propose a hierarchical dataflow-driven architecture to accelerate WBP. A pack-and-ship approach is presented under a non-uniform memory access (NUMA) architecture to allow the subordinate tiles to operate in a bundled access and execute manner. We also propose a multi-level dataflow model and the related scheduling scheme to manage and allocate the heterogeneous hardware resources. Experiment results demonstrate that our prototype achieves $2\times$ and $2.3\times$ speedup in terms of normalized throughput and single-tile clock cycles compared with GPU and DSP counterparts in several critical WBP benchmarks. Additionally, a link-level throughput of $288$ Mbps can be achieved with a $45$-core configuration.