PaperlayerView exact revision on arXiv

Source-linked AI summary

Sensing Assisted Satellite Backhaul with FBL UL and Broadcast DL for Massive IoT

Tijana Devaja, Milica Petkovic, Israel Leyva-Mayorga, Dejan Vukobratovic, Cedomir Stefanovic

arXiv:2609.28588v1cs.ITmath.IT

TL;DR

Massive IoT short packets face finite-blocklength reliability limits, while satellite backhaul for remote deployments is vulnerable to atmospheric attenuation and prior studies often treat network components separately. This paper jointly models uplink random access, sensing-assisted satellite backhaul, and worst-user broadcast downlink. Its numerical results show an optimal uplink access probability and improved backhaul robustness from sensing margins, with a throughput tradeoff.

  • Problem

    Massive IoT uplink, satellite backhaul, and downlink reliability have largely been studied separately, despite short-packet finite-blocklength effects and attenuation-sensitive satellite links.

    Source: §I · p. 1

  • Method

    The paper unifies finite-blocklength uplink random access, sensing-assisted satellite backhaul adaptation, and worst-user broadcast downlink in an end-to-end model.

    Source: §I · p. 1

  • Results

    The analysis finds an optimal uplink access probability for sum rate, improved backhaul robustness from sensing margins at reduced throughput, and a broadcast downlink governed by worst-user SNR.

    Source: §V · p. 6

  • Takeaways & Limitations

    Coordinating uplink access, sensing margins, and downlink blocklength and rate addresses the reported end-to-end reliability and bottleneck concerns.

    Source: §V · p. 6

Abstract

from arXiv · show

Massive Internet of Things (IoT) networks operate with short packets whose reliability is limited by finite block- length (FBL) effects, while remote deployments increasingly rely on Low Earth Orbit (LEO) satellites for backhaul connectivity that is sensitive to atmospheric attenuation. In this paper, we pro- pose a unified end-to-end framework for satellite-assisted massive IoT networks that jointly models uplink FBL random access, sensing assisted satellite backhaul, and worst user broadcast downlink transmission. Uplink reliability is characterized using stochastic geometry, while atmospheric sensing and conservative SNR margins enable FBL-safe backhaul adaptation. Numerical results reveal an optimal uplink access probability due to the tradeoff between spatial reuse and FBL reliability, and show that sensing assisted backhaul margins significantly improve robustness against attenuation uncertainty.

I. INTRODUCTION

Massive IoT couples short-packet uplink reliability with satellite backhaul vulnerable to atmospheric attenuation, but prior work generally treats uplink reliability, backhaul adaptation, and downlink robustness separately. This paper unifies these components under finite-blocklength constraints and reports an optimal uplink access probability and more robust backhaul adaptation using sensing margins.

I. INTRODUCTION

  • Prior work largely studied uplink reliability, satellite backhaul adaptation, and downlink robustness separately, leaving their end-to-end interaction unmodeled.The cited gap includes uplink evaluations assuming ideal backhaul and satellite optimization based on asymptotic rates without uplink-induced traffic load.Source: §I · p. 1
  • The framework jointly models stochastic-geometry uplink random access, sensing-assisted LEO satellite backhaul, and worst-user broadcast downlink under finite-blocklength constraints.Uplink sensing data are forwarded through satellite backhaul; control information and feedback return to devices by broadcast.Source: §I · p. 1
  • Sensing-based attenuation estimates support conservative effective-SNR calculation and finite-blocklength-safe per-hop backhaul rates.The approach is framed as consistent with integrated sensing and communications principles.Source: §I · p. 1
  • An optimal uplink access probability maximizes successful sum rate, while sensing margins mitigate attenuation uncertainty and worst-user SNR can bottleneck broadcast downlink.Raw sensing-driven SNR-based rate selection can cause mismatch and outages; the paper also presents downlink error probability within end-to-end reliability.Source: §I · pp. 1–2

II. SYSTEM MODEL

The system model treats all communication phases as short-packet transmissions carrying a common payload, with distinct blocklengths and finite-blocklength reliability across uplink, satellite backhaul, and downlink. Its network-topology model places devices and base stations as independent Poisson point processes and associates each device with its nearest base station.

A. Network Topology

  • Devices and base stations are independently distributed as homogeneous Poisson point processes with densities λu and λb, respectively.Each device associates with its nearest base station.Source: A. Network Topology · p. 2
  • The model assumes normalized unit transmit power, Nakagami fading, path loss r^-η with η > 2, and an interference-limited regime.The Nakagami parameter m = 1 corresponds to Rayleigh fading.Source: A. Network Topology · p. 2

II. SYSTEM MODEL

  • The model uses a common k-bit payload across uplink, backhaul, and downlink, while allowing a different blocklength for each phase.Packets are represented with a finite-blocklength reliability abstraction across all hops, including conservative rate selection under imperfect channel knowledge.Source: A. Network Topology · p. 2, §II · p. 2

B. UL IoT-to-BS Random Access

The uplink random-access model represents independently active devices as interferers at a reference base station, then characterizes interference and the typical active device’s SINR. It uses the interference Laplace transform and specializes part of the analysis to path-loss exponent η = 4 for tractability.

B. UL IoT-to-BS Random Access

  • Each device transmits independently in a slot with probability p, and the reference base station’s aggregate interference sums contributions from active devices other than the typical device.Each contribution depends on the interfering device’s fading power and distance to the reference base station.Source: B. UL IoT-to-BS Random Access · p. 2
  • The typical active device’s uplink SINR depends on its distance and fading to its serving base station and the aggregate interference, with background noise negligible in the stated regime.The model treats distance, fading, and interference as statistically independent under the stated Poisson and independent-fading assumptions.Source: B. UL IoT-to-BS Random Access · p. 2
  • For Poisson-distributed devices, ALOHA activity, and Nakagami-m fading, the interference distribution is characterized through its Laplace transform.For analytical tractability, part of the analysis specializes to η = 4, where the interference density is a Lévy distribution.Source: B. UL IoT-to-BS Random Access · pp. 2–3

C. Satellite Backhaul and Sensing Model

The satellite backhaul model accounts for rain attenuation and pointing loss on both satellite-link directions and uses attenuation sensing to form SNR estimates. Because those estimates may be optimistic, conservative SNR margins are introduced to reduce finite-blocklength rate-selection mismatch, trading achievable rate against robustness.

C. Satellite Backhaul and Sensing Model

  • Rain attenuation and pointing loss enter the uplink base-station-to-satellite and downlink satellite-to-base-station SNRs, alongside link gains, path loss, transmit power, and receiver noise.The attenuation factor is modeled as at least one in Ka/Ku bands.Source: C. Satellite Backhaul and Sensing Model · p. 3
  • Satellite sensing estimates rain attenuation in the logarithmic domain, with Gaussian sensing uncertainty characterized by σ.The estimates are mapped to linear scale to form corresponding estimated SNRs.Source: C. Satellite Backhaul and Sensing Model · p. 3
  • Log-domain estimation followed by linear mapping generally yields biased, optimistic SNR estimates, while finite-blocklength performance is sensitive to SNR uncertainty and lower-tail events.The stated motivation is to address uncertainty rather than only correct mean estimation error.Source: C. Satellite Backhaul and Sensing Model · p. 3
  • Conservative uplink and downlink SNR margins reduce sensing-error rate mismatches, but residual outage may remain and the margins trade achievable rate against robustness.A mismatch occurs when the selected coding rate exceeds instantaneous channel support.Source: C. Satellite Backhaul and Sensing Model · p. 3

D. Broadcast DL Model

The model considers sensing information delivered to multiple subscribers, making broadcast a resource-efficient alternative to per-user unicast. Each cell uses one rate chosen to meet a target error probability at its worst-SNR device.

D. Broadcast DL Model

  • Broadcasting one message to multiple subscribers is presented as a resource-efficient alternative to per-user unicast, with reliability constrained by the worst user.The service model covers sensing information forwarded to a destination cell and disseminated to users subscribing to the same service.Source: D. Broadcast DL Model · p. 3
  • Each base station sends a common downlink packet per slot and selects one code rate per cell rather than adapting the rate per user.Examples include a control message or aggregated status.Source: D. Broadcast DL Model · p. 3
  • The selected cell rate must meet the target error probability at the device with the worst downlink SNR.The passage identifies the cell-edge or a low-percentile user as a practical example of a conservative SNR choice.Source: D. Broadcast DL Model · p. 3

III. FINITE-BLOCKLENGTH PERFORMANCE ANALYSIS

The analysis defines packet reliability under FBL constraints and develops an uplink block-error probability for short packets. The resulting approximation captures system-parameter effects and supports performance analysis.

III. FINITE-BLOCKLENGTH PERFORMANCE ANALYSIS

  • Reliability is the probability of successful packet decoding under FBL constraints; uplink reliability concerns decoding at the serving base station.For backhaul and broadcast downlink, target block error probabilities constrain the maximum supported transmission rates.Source: §III · p. 3

A. UL Block Error Probability

  • The uplink block-error analysis uses the FBL normal approximation for a given SINR, code rate, blocklength, and target packet error probability.The approximation uses Shannon capacity and channel dispersion, and the unconditional uplink error probability is derived using SINR.Source: A. UL Block Error Probability · pp. 3–4
  • The short-packet uplink error probability is tightly approximated using a linear approximation of the Q-function because the SINR integral has no closed form.The resulting expression captures path loss, fading severity, and blocklength effects and can be evaluated for performance analysis.Source: A. UL Block Error Probability · p. 4

B. Aggregate UL Throughput and Sum Rate

The section defines uplink throughput from the mean number of successfully decoded packets and delivered information bits. Its goodput metric accounts for random access, interference, and FBL decoding errors.

B. Aggregate UL Throughput and Sum Rate

  • The mean number of successfully decoded uplink packets per slot depends on device density in the cell, access probability, and packet decoding success probability.Decoding success is expressed as 1 − P_e^ul, with P_e^ul given by the uplink error analysis.Source: B. Aggregate UL Throughput and Sum Rate · p. 4
  • The uplink sum rate is defined as the mean number of successfully delivered information bits per slot in a cell.The section also gives an equivalent spectral-efficiency form.Source: B. Aggregate UL Throughput and Sum Rate · p. 4
  • The resulting mean uplink goodput explicitly accounts for random access, interference, and FBL decoding errors, unlike asymptotic Shannon sum rates.It reflects reliably delivered information bits per slot.Source: B. Aggregate UL Throughput and Sum Rate · p. 4

C. Backhaul FBL-Constrained Rates

Backhaul rates are constrained by FBL reliability, with hop rates derived from effective SNRs and the end-to-end usable rate set by the bottleneck hop. Allocated time-bandwidth resources determine the information bits delivered in a slot.

C. Backhaul FBL-Constrained Rates

  • The FBL normal approximation gives the maximum reliable backhaul rate for a given SNR, blocklength, and target block error probability.The backhaul blocklength is n_bh and its target block error probability is ϵ_bh.Source: C. Backhaul FBL-Constrained Rates · p. 4
  • Applying the FBL rate relation to the effective SNRs of both hops yields FBL-safe uplink and downlink rates.When nominal hop rates are specified, the actually usable rates are clipped; the two-hop end-to-end usable rate is the bottleneck value.Source: C. Backhaul FBL-Constrained Rates · p. 4
  • The information bits delivered end-to-end in a slot depend on the satellite bandwidth, slot duration, and the fraction of time-bandwidth resource allocated to the base station.The passage relates delivered bits to allocated resource and achievable coding rate.Source: C. Backhaul FBL-Constrained Rates · p. 4

D. Broadcast FBL Reliability and End-to-End Success

The broadcast downlink rate is constrained by its target FBL error probability at the edge user. End-to-end success combines uplink decoding, two-hop backhaul, and broadcast reception, under independence and no-queueing assumptions.

Broadcast FBL reliability

  • The admissible broadcast rate is constrained by the target block-error probability at the edge user across cells.The edge-user error probability is evaluated for the selected downlink blocklength and rate.Source: D. Broadcast FBL Reliability and… · p. 4

End-to-end success

  • End-to-end packet success combines uplink decoding, two-hop backhaul success, and broadcast downlink reception.The approximation assumes independence among these events and neglects queueing-induced packet drops and excessive delays.Source: D. Broadcast FBL Reliability and… · pp. 4–5
  • The accompanying uplink plot shows sum rate versus access probability for several code rates.Source: Figure 2 · p. 5
  • Improvements in uplink FBL reliability, backhaul effective SNR, or broadcast SNR translate into higher end-to-end reliability.Source: D. Broadcast FBL Reliability and… · p. 5

IV. NUMERICAL RESULTS

Numerical results show an uplink access-probability tradeoff, robustness gains from conservative sensing margins, and improved end-to-end reliability with longer uplink blocklengths. The evaluation assumes similar rain attenuation on the backhaul’s forward and return links.

Numerical results

  • The evaluation assumes identical rain attenuation in the uplink and downlink backhaul directions to isolate sensing uncertainty and FBL adaptation.The assumption is considered reasonable for links sharing a propagation path and nearby frequency bands.Source: §IV · p. 5
  • The uplink sum rate peaks and then collapses as access probability rises, reflecting the tradeoff between spatial reuse and FBL decoding reliability.More simultaneous transmissions increase spatial reuse but also aggregate interference and uplink error probability.Source: §IV · p. 5
  • Fig. 3 shows usable backhaul rate falling with rain attenuation; conservative sensing margins narrow performance spread and improve robustness, with the mean-rate threshold at 11 dB.The threshold marks where the mean backhaul rate meets the required rate; margins reduce achievable rate while improving robustness to estimation errors.Source: Figure 3 · p. 5, §IV · pp. 5–6
  • As device density increases, end-to-end success probability decreases, while longer uplink blocklengths improve reliability, particularly at moderate and high densities.The figure fixes access probability at 0.05 and payload at 32 bits; backhaul and broadcast downlink error probabilities are fixed at 10^-3 and 10^-4.Source: Figure 4 · p. 6, §IV · p. 6

V. CONCLUSION

The conclusion presents a unified FBL model of uplink random access, sensing-assisted satellite backhaul, and worst-user broadcast downlink. It highlights an optimal uplink access probability and a robustness-throughput tradeoff for backhaul sensing margins.

V. CONCLUSION

  • The unified end-to-end model jointly captures uplink FBL random access, sensing-assisted satellite backhaul, and worst-user broadcast downlink.Source: §V · p. 6
  • The tradeoff between uplink spatial reuse and FBL reliability yields an access probability that maximizes uplink sum rate.Source: §V · p. 6
  • Conservative backhaul sensing margins improve robustness to attenuation estimation errors at the cost of reduced throughput.Broadcast downlink performance is governed by the worst-user SNR, so blocklength and rate selection matter to avoid an end-to-end bottleneck.Source: §V · p. 6
Loading 2609.28588v1…