Accurate prediction of the future trajectories of surrounding agents is essential for safely autonomous vehicles. However, it is quite challenging due to the dynamics of driving environments caused by frequent agent-agent and agent-map interactions. Most existing methods mainly focus on modeling the dynamics of agents but overlook the inherent dynamics of the usable map at different driving moments. To address this limitation, this paper proposes a bidirectional interaction network (DyMap) that takes into account the bilateral constraints of the map on the agents and the impact of the agents on the map, using a unique bidirectional attention module. By employing multiple stages of bidirectional interaction modules, the features of agents and map nodes are continuously updated to adapt to changing environments. To facilitate network learning, we further incorporate map-related auxiliary tasks into the learning objectives, including traffic flow prediction, map node occupancy prediction, and road direction prediction. These tasks are designed to help the network learn map features that can reflect the dynamics of the driving environment, leading to a comprehensive feature representation of the agents. This representation encodes not only the agent’s historical motions but also the latent map information triggered by other agents. This proposed network offers a new approach to modeling environmental dynamics in driving scenarios, and achieves the state-of-the-art performance on the Argoverse 1 and Argoverse 2 trajectory prediction benchmarks. Note to Practitioners—Autonomous vehicles are of particular interest to practitioners in the field of automation science and technology. Trajectory prediction is of critical importance to the safety of autonomous driving, which aims to forecast the future positions of a given traffic participator (called the agent in this paper, such as a vehicle, pedestrian, etc.) based on the observations of itself and other agents’ historical trajectories as well as the map information. This paper proposes a trajectory prediction network based on a novel bidirectional interaction module that enables bidirectional information flow between the features of map nodes and agents so as to simultaneously learn their feature representations being aware of the timing-vary property of traffic scenes. The proposed DyMap contains three blocks of bidirectional interaction modules, which are interleaved with self-interactions of the agents and map nodes. Three different kinds of map-related tasks are developed to facilitate network training by leveraging on maximizing the map features for predicting traffic flow, occupancy, and lane direction. The superiority of the proposed method has been confirmed by the widely used benchmarks in the community. Besides modelling the complex interaction between agents and map nodes in dynamic traffic scenarios, the proposed bidirectional interaction network can be beneficial to address other problems requiring symmetric modelling of interactions between different components, such as human-robot interaction. The proposed map-related tasks can be directly used in other applications requiring to extract map features while considering the time-varying conditions, including but not limited to robot navigation and embody intelligence.

The possibility for public attributes to disclose private information has caused widespread concern. Traditional privacy-preserving approaches have two limitations: 1) Approaches based on data anonymization or distortion often lead to poor utility-privacy trade-offs, and 2) approaches based on data encryption face heavy computational costs. These problems have prompted calls for an effective privacy-preserving framework that provides adequate privacy guarantees while maintaining good data utility. Inspired by denoising autoencoders, in this paper, we regard the information about privacy attributes contained in the public attributes as a kind of noise and design an ex ante privacy-preserving model called the Mutual Information Autoencoder (MIAE), which reconstructs the loss function of the original autoencoder by combining reconstruction errors and mutual information, and we introduce a trade-off coefficient to achieve utility-privacy trade-offs. To elucidate the superiority of the proposed model, we consider utility-privacy trade-offs with the expected distortion function as a metric of data utility and the joint mutual information as a metric of privacy disclosure, and then, we construct a convex optimization problem with multiple constraints based on rate-distortion theory. From an information theory perspective, we provide a lower bound for privacy disclosure with utility guarantees. Elaborate experiments over a real-world dataset reveal that as the level of expected distortion increases, the achievable bound obtained by MIAE exhibits a trend similar to that of the information-theoretic bound. When the expected distortion surpasses 2.2, the achievable bound obtained by MIAE also converges to 0, and the maximum gap between the achievable bound obtained by MIAE and the information-theoretic bound is no more than 1.4. Compared to existing models, MIAE can provide a tighter achievable bound and achieve good utility-privacy trade-offs.

We consider frequency-flat time-varying fading channels with no channel model information (CMI). By introducing the smoothness function to measure the extent of channel fluctuation, we derive a robust soft-decision-aided (SDA) channel estimator based on Pareto optimality of the double-objective optimization of the likelihood function of the received signal sequence and the smoothness constraint of the channel estimates. Compared with the conventional maximum-likelihood-based channel estimators derived under the block-fading assumption, the newly derived SDA-Pareto estimator gives more freedom to the channel estimation process, allowing channel estimates to have controlled variations to track the time-varying channel more closely. Compared with estimators derived based on the maximum a posteriori probability or the minimum mean-square error criterion which require explicit acquisition of the CMI, the SDA-Pareto estimator significantly simplifies the channel measurement process by requiring only a suitable regularization parameter to balance the trade-off between the likelihood function and the smoothness condition. An adaptive algorithm is proposed to adjust the regularization parameter adaptively, enabling an efficient and effective implementation of the SDA-Pareto estimator in practical applications. Simulation studies are provided to demonstrate the advantage of the SDA-Pareto estimator over the conventional estimators in both channel estimation accuracy and error-rate performance.

We consider iterative decoding of repeat- accumulate (RA) codes over frequency-flat, quasi- static fading channels. A soft-input, soft-output decoder is proposed for the inner convolutional decoding, which fuses the decoding approach of the soft-output Viterbi algorithm and the estimation approach of the maximum-likelihood sequence detector. The decoder deploys trellis search algorithm based on the generalized likelihood ratio test, whereby the channel state information is acquired implicitly using both the pilot and data signals during the decoding process. Through simulations, we show that the RA decoding with the proposed decoder has much better error performance than standard RA decoding with pilot-symbol- assisted channel estimation, while having approximately the same computational complexity. Compared with the conventional scheme of iterative channel estimation and decoding, the proposed decoder has much simpler structure and requires significantly less computational power, although it incurs some loss in error performance.

Accurate channel state information (CSI) is crucial for reliable decoding over fading channels. In many existing works, the CSI acquisition relies greatly on accurate knowledge of the channel model information (CMI), which includes, in particular, the power spectrum of the fading process and the statistical distribution of the fading gain. In practice, it may be difficult to obtain or keep track of the CMI accurately. In this paper, we consider the decoding of Reed- Solomon convolutional concatenated (RSCC) codes over the quasi-static fading channel, and propose a robust, decision-directed, soft-output convolutional decoder which does not require explicit CSI acquisition. Simulation result shows that the proposed decoder has a lower bit-error rate for convolutional decoding than the conventional soft-output Viterbi algorithm and the BCJR algorithm with pilot-symbol-assisted channel estimation. The advantage of the former in terms of the accuracy of soft-decision output is demonstrated by the superiority of word-error performance in the RSCC decoding.

We consider the differential phase-shift keying modulation over the flat Rician fading channel with perfect channel model information (PCMI), and derive the two-symbol-observation-interval log-likelihood ratio (LLR) metric from first principles. The work generalizes the previous LLR derivations in [16]-[19], and demonstrates the necessity of the knowledge of the statistical channel model in the exact LLR computation. Assuming the channel is slowly time-varying, we also propose a simple approximate LLR metric based on the generalized likelihood ratio test (GLRT), which requires only the information of the noise spectral density. The advantage of the newly derived PCMI-LLR metric over the approximate metric and the other LLR metrics in the literature is explained from the information-theoretic perspective via the generalized mutual information. Computer simulations are also provided to demonstrate the superiority of the PCMI-LLR metric in both error rate performance and convergence speed for iterative decoding of turbo and low-density parity-check codes in practical systems. This paper aims to present the fundamental principles of LLR computation, emphasizing the importance of selecting the appropriate metric for iterative decoding over fading channels.

We consider transmissions over the frequency-nonselective time-selective fading channel and derive the soft-decision-aided maximum a posteriori probability (SDA-MAP) channel estimator that utilizes the entire received signal sequence, including both pilot signals and data signals, and the a priori statistical information of the transmitted data symbols. This paper theoretically demonstrates how the soft data decisions should be incorporated into the channel estimation process. An iterative implementation of the SDA-MAP estimator is proposed, which uses the feedback of earlier soft decisions and the feedforward of later soft decisions to estimate the current channel gain. This approximate SDA-MAP implementation is similar in structure to the expectation-maximization algorithm. Simulation results show that the receiver using this iterative SDA-MAP estimator outperforms the conventional pilot-symbol-assisted modulation receiver in which the channel estimates are obtained from only the pilot signals. An application to iterative decoding using the LTE turbo code is also provided.

We consider the iterative decoding of low-density parity-check codes with M-ary phase-shift keying transmission and pilot-symbol-assisted channel estimation over time correlated Rayleigh fading channels. We present the correct conceptual approach for deriving the log-likelihood ratio (LLR) expression for a general q-ary code. Its bit-error probability (BEP) performance is compared with that of the conventional metric which does not take into account the information concerning the channel estimation accuracy. Simulation results show that this LLR metric outperforms the conventional metric in both BEP performances and average number of decoding iterations required for convergence. The work here demonstrates the importance of incorporating the knowledge of the channel estimation accuracy in the iterative decoding process.

We derive the exact log-likelihood ratio (LLR) for iterative decoding of low-density parity-check codes for pilot-symbol-assisted BPSK transmission over flat Rayleigh fading channels. The bit-error rate (BER) performances of this metric are studied for various interleavers and different pilot spacings. We further compare this metric with the approximate metric, which has been commonly used in the literature. Simulation results show that the newly derived LLR metric results in better BER performances. In addition, the effect of signal-to-noise ratio estimation error is investigated and our new LLR shows more robustness.