Source-linked AI summary
Partial Superimposed Pilot-Aided Sparse Vector Transmission for High-Mobility URLLC
Yanfeng Zhang, Jinkai Zheng, Hui Liang, Wen Wu, Tom H. Luan, Di Zhang, Xu Zhu
TL;DR
Rapid within-packet channel variation in high-mobility scenarios makes accurate channel acquisition challenging for reliable decoding. PSP-SVT uses a few pilots on selected subcarriers with an iterative joint channel-estimation and data-decoding receiver; simulations report better BLER than existing full-SP-aided SVT schemes with fast convergence.
Problem
In high-mobility scenarios, rapid channel variation within a packet makes accurate channel acquisition challenging and poses a bottleneck for reliable decoding.
Source: §I · p. 1Method
PSP-SVT places a limited, uniformly distributed set of pilots on a subset of subcarriers and uses an iterative receiver for joint channel estimation and data decoding.
Source: §I · p. 2, §II · p. 2Results
PSP-SVT outperforms existing full-SP-aided SVT schemes in BLER and typically reaches stable performance within two iterations.
Source: §IV · pp. 5–6, §V · p. 6Takeaways & Limitations
The tested PSP-SVT configuration has an optimal data power allocation ratio of about 0.7 versus 0.63 for SSVT, while maintaining reliable channel estimation.
Source: §IV · p. 4
Abstract
from arXiv · showhide
A partial superimposed pilot-aided sparse vector transmission (PSP-SVT) scheme is proposed for short-packet ultra-reliable and low-latency communications in high-mobility scenarios. Unlike conventional full superimposed pilot-aided SVT schemes, the proposed PSP-SVT scheme deploys only a few pilots over a subset of subcarriers. This sparse pilot structure is sufficient for basis expansion model based channel tracking while effectively reducing pilot-data interference. Based on the PSP pattern, an iterative receiver is developed to jointly perform channel estimation and data decoding. The reduced pilot interference in PSP-SVT provides more accurate initial channel estimation and data detection, thereby improving the subsequent iterative refinement and mitigating their error propagation. Moreover, the impacts of the number of PSPs and power allocation ratio on block error rate (BLER) performance are investigated to reveal the near-optimal pilot configuration. Simulation results show that the proposed PSP-SVT scheme outperforms existing full superimposed pilot-aided SVT schemes in terms of BLER with fast convergence speed.
I. Introduction
Short packets leave limited room for pilots, while high mobility makes channel acquisition and reliable decoding difficult. PSP-SVT addresses the interference from full superimposed pilots with a sparse pilot pattern and an iterative receiver.
I. Introduction
- Short data packets leave limited room for pilot overhead and decoding latency; in high-mobility environments, rapid channel variation within a packet duration imposes additional challenges on reliable transmission.Source: §I · p. 1
- Full-SP SSVT and JPD-SVT suffer pilot–data interference that limits initial channel estimation and data decoding, constraining later iterative decoding.JPD-SVT reuses reliably decoded symbols as pseudo-pilots, but still relies on full SPs.Source: §I · pp. 1–2
- PSP-SVT deploys a small number of pilots over only part of the subcarriers, aligning pilot placement with the low-dimensional BEM channel representation.This pattern reduces pilot–data interference, channel-estimation complexity, and supports reliable initialization for iterative decoding.Source: §I · p. 2
- The proposed iterative receiver uses reduced pilot interference for more accurate initial BEM estimation, mitigating error propagation; simulations report superior BLER to existing SVT schemes.The study also investigates the number of PSPs and power allocation ratio to identify a near-optimal pilot configuration.Source: §I · p. 2
II. Encoding Process of PSP-SVT Scheme
The encoding process maps information into a sparse data vector, spreads it with a codebook, and superimposes pilots on selected subcarriers. Slepian-BEM modeling reduces the channel-estimation parameter count, motivating an iterative receiver that jointly estimates the channel and decodes data.
II. Encoding Process of PSP-SVT Scheme
- The data vector carries index information in its active support and symbol information in its nonzero entries, then codebook spreading produces the coded data signal.The transmission block uses N_f subcarriers, and the sparse vector has K nonzero entries.Source: §II · p. 2
- The PSP vector places N_P uniformly spaced pilots on a subset of subcarriers, avoiding clustered observations and excessive overlap of pilot-induced regions.Pilot spacing is N_f/N_P.Source: §II · p. 2
- Slepian-BEM models each time-varying channel path compactly, reducing the unknown channel parameters from N_fL to QL.In finite-scattering environments, QL is typically much smaller than N_f, and at least N_P ≥ QL pilots are sufficient to estimate the BEM coefficients.Source: §II · p. 3
- The channel model expresses the time-domain channel matrix through Slepian basis functions and coefficients, with W representing modeling error.The frequency-domain formulation uses F_L and the coefficient vectors c_q to express the channel model.Source: §II · p. 3
- The receiver first estimates the channel from the PSP signal and decodes data, then reuses decoded symbols as pseudo-pilots for channel refinement and further decoding.The joint iterative algorithm is developed based on the received-signal model.Source: §III · p. 3, §II · p. 3
A. Initial Channel Estimation and Data Decoding
The receiver uses the PSP pattern’s low-interference subcarriers to initialize channel estimation and data decoding. It estimates the BEM coefficients, reconstructs the channel, removes the pilot signal, and decodes the data.
A. Initial Channel Estimation and Data Decoding
- PSP energy is concentrated within Q neighboring subcarriers, leaving low-interference regions for more accurate initial channel estimation and data decoding.The receiver collects the Q neighboring received signals around each PSP position to form its initial estimation model.Source: A. Initial Channel Estimation and Data… · p. 3
- The receiver extracts signals indexed by S to form an initial channel-estimation model with a BEM coefficient vector and measurement matrix.S is the union of local received regions associated with the PSP positions.Source: A. Initial Channel Estimation and Data… · p. 4
- The initial BEM coefficient estimate is obtained from the signal model using the selected received signals.The estimate is formed at iteration 0.Source: A. Initial Channel Estimation and Data… · p. 4
- The receiver reconstructs the initial frequency-domain channel matrix and removes the received pilot signal to obtain the data signal.The reconstructed channel is used for initial data decoding.Source: A. Initial Channel Estimation and Data… · p. 4
- Initial data decoding recovers the sparse vector with MMP and demodulates the detected support’s QAM symbols.The MMP algorithm is used to solve the sparse-vector recovery problem.Source: A. Initial Channel Estimation and Data… · p. 4
B. Iterative Enhanced Data Decoding
The iterative receiver uses decoded data from the previous iteration as pseudo-pilots to refine channel estimation and decoding. The PSP pattern’s localized pilot energy supports this refinement and is reported to improve convergence and robustness over full-SP SVT schemes.
B. Iterative Enhanced Data Decoding
- At each iteration, the receiver uses the previous iteration’s decoded sparse data vector as pseudo-pilots and combines its reconstructed spread signal with the original PSP.This forms a new pilot signal for channel estimation.Source: B. Iterative Enhanced Data Decoding · p. 4
- The new pilot signal forms an updated channel-estimation model, from which the receiver updates BEM coefficients and reconstructs the frequency-domain channel matrix.The model uses the q-indexed submatrices of the measurement matrix.Source: B. Iterative Enhanced Data Decoding · p. 4
- The receiver decodes data using the updated channel and repeats channel estimation and decoding until it reaches the maximum iteration count or the convergence criterion.Source: B. Iterative Enhanced Data Decoding · p. 4
- The PSP pattern’s Q-subcarrier pilot-energy concentration creates low-interference regions that support pseudo-pilots, suppress error propagation, and improve convergence and robustness over full-SP SVT schemes.The decoded data symbols serve as pseudo-pilots in subsequent iterations.Source: B. Iterative Enhanced Data Decoding · p. 4
IV. Simulation Results
Simulations show that PSP-SVT achieves better BLER than the compared pilot-aided SVT schemes, while pilot count and power allocation shape its performance. The scheme also converges quickly, with its advantage becoming more pronounced at larger subcarrier counts.
IV. Simulation Results
- The BLER-optimal data power allocation ratio is about 0.7 for PSP-SVT versus 0.63 for SSVT.BLER first decreases and then increases with αd; PSP-SVT uses 14 PSPs in this comparison.Source: §IV · p. 4, Figure 3 · p. 5
- BLER degrades sharply when NP < 6, while adding PSPs beyond sufficient BEM observations offers limited gain and can worsen interference.NP = 14 is adopted for subsequent simulations as a favorable BLER tradeoff across SNRs.Source: Figure 4 · p. 5, §IV · p. 5
- PSP-SVT achieves the best BLER at different SNR values compared with SSVT and JPD-SVT, using 14 pilots versus their 84.The comparison uses K = 2; the PSP-SVT pilot pattern reduces pilot–data interference and improves receiver initialization.Source: Figure 5 · p. 5, §IV · p. 5
- PSP-SVT outperforms JPD-SVT for all tested Nf values, with the performance gain becoming more pronounced as Nf grows.With a fixed number of pilots, larger Nf makes deployment sparser and enlarges low-interference regions.Source: Figure 6 · p. 5, §IV · p. 5
- PSP-SVT has lower initial BLER than SSVT for K = 2 and K = 3 and typically reaches stable performance within two iterations.The receiver burden mainly comes from BEM-based channel estimation and sparse data decoding at each iteration.Source: §IV · pp. 5–6, Figure 7 · p. 6
V. Conclusions
The conclusion presents PSP-SVT for short-packet URLLC over high-mobility channels, using sparse pilots and an iterative receiver. Simulations report better BLER than existing full-SP-aided SVT schemes and fast convergence.
V. Conclusions
- PSP-SVT targets short-packet URLLC over high-mobility channels with a small number of pilots on a subset of subcarriers.The sparse pattern reduces pilot–data interference and creates low-interference regions for receiver initialization.Source: §V · p. 6
- The iterative receiver jointly estimates the channel and decodes data under Slepian-BEM modeling, reusing decoded symbols as pseudo-pilots.This supports iterative refinement.Source: §V · p. 6
- Simulations show better BLER than existing full-SP-aided SVT schemes, with fast convergence.Source: §V · p. 6