PaperlayerView exact revision on arXiv

Source-linked AI summary

S-PASS: Star-Topology Pinching-Antenna Systems with Pipelined Antenna Pre-Positioning

Wei Jiang, Hans D. Schotten

arXiv:2610.29700v1cs.ITeess.SPmath.IT

TL;DR

Mechanical re-pinch latency challenges PASS’s per-user placement advantage, which assumes repositioning takes no time. S-PASS combines star-connected waveguide arms on one RF chain with PAP, overlapping idle-arm antenna moves with transmission; under balanced loads and equal move times, PAP’s latency threshold for matching two-user NOMA rises from 0.0067 slot durations with one arm to 7.05 with eight.

  • Problem

    PASS’s per-user re-pinch advantage assumes repositioning takes no time, although antenna reconfiguration can take milliseconds to seconds.

    Source: §I · p. 1, A. Re-pinching Latency Model · p. 4

  • Method

    S-PASS connects radial waveguide arms to one RF chain through a switch, while PAP moves idle-arm antennas to their next users as another arm transmits.

    Source: §I · p. 1

  • Results

    0.0067 slot durations with one arm and 7.05 with eight are the latency limits at which PAP at least matches two-user NOMA.

    Source: §VI · p. 6

  • Takeaways & Limitations

    With users directly below the arms after latency is absorbed, the outer-endpoint user gains 17% over fixed-placement TDMA, compared with 6.7% in expected sum rate.

    Source: §VI · p. 6

    The reported scheduling and latency results assume balanced loads and equal move times.

    Source: §VI · p. 6

Abstract

from arXiv · show

Pinching-antenna systems (PASS) flexibly reconfigure the wireless channel by positioning dielectric antennas along a waveguide near each user. Releasing and re-attaching an antenna, however, is a mechanical operation orders of magnitude slower than electronic switching, wasting much transmission time. Existing designs generally assume instant reconfiguration, neglect the mechanical latency, or keep the antenna fixed. This paper proposes S-PASS, a star architecture of waveguides sharing one radio-frequency chain through a switch, and pipelined antenna pre-positioning (PAP), a time-division protocol in which idle waveguides pre-position their antennas for their next users while one transmits. It derives in closed form the scheduling efficiency, the number of waveguides beyond which the latency costs no transmission time, and the latency up to which PAP outperforms fixed placement. Analytical and numerical results show that this latency grows with each added waveguide, and users at the outer ends of the waveguides gain most.

I. INTRODUCTION

Re-pinch latency can leave a PASS channel idle for many slots, while fixed placement avoids that overhead but can penalize distant users. S-PASS pairs a switched star of waveguide arms with pipelined antenna pre-positioning (PAP), which overlaps idle-arm movement with transmission and analyzes when the latency is absorbed.

I. INTRODUCTION

  • Earlier approaches model electronic switching, add a guard interval, or allow antenna roaming, but none lets an idle PA move while another waveguide transmits.Re-pinch movement can take milliseconds to seconds, whereas scheduling is much faster; fixed-placement NOMA avoids the movement overhead.Source: §I · p. 1
  • PAP lets idle-arm antennas move toward their next users while another arm transmits, absorbing latency completely once enough arms share a balanced load and partly otherwise.The analysis assumes quasi-static users, balanced arm loads, and equal movement times for the interleaved schedule.Source: §I · pp. 1–2
  • S-PASS connects several radial waveguide arms to one RF chain through a central switch, with one movable PA per arm.A single PA per arm leaves the uplink free of inter-antenna radiation that couples multiple active pinches on one waveguide.Source: §I · p. 1
  • The paper derives closed-form PAP scheduling efficiency, a sufficient arm count for eliminating latency costs, and latency thresholds against fixed-placement baselines.The baselines include fixed-placement NOMA and fixed-placement TDMA.Source: §I · pp. 1–2
  • Each added arm increases PAP’s tolerable latency by the same amount for a fixed number of users per arm and a given fixed-placement baseline.In the evaluated freespace setting, re-pinching improves fifth-percentile and mean spectral efficiency over fixed-placement TDMA once enough arms share the load.Source: §I · p. 2

II. SYSTEM MODEL

S-PASS uses a central hub and switch to connect one RF chain to radial waveguide arms, each serving users in its nearest-arm sector. Its channel analysis assumes line-of-sight freespace propagation and neglects in-guide and switch losses.

A. Star Architecture

  • One RF chain at the central hub connects through a 1:S electronic switch to separate radial waveguide arms, each with its own feed and movable PA.The arms partition the coverage disk into nearest-arm sectors; their radial placement problems are independent.Source: A. Star Architecture · p. 2
  • Each user is assigned to its nearest-arm sector, and the PA’s feasible target is the user’s projection onto that arm.The analysis sets the lateral offset to zero, while Fig. 4 relaxes that assumption.Source: A. Star Architecture · p. 2

B. Channel and Signal Model

  • The rate analysis uses a line-of-sight freespace model for mmWave and neglects in-guide loss and switch insertion loss.The terahertz regime, where in-guide loss can be substantial, is left to future work.Source: B. Channel and Signal Model · p. 2

1) Downlink:

Downlink rate depends on the PA–user distance and is maximized when the PA is directly over the user. In the uplink, one PA per arm avoids the inter-antenna radiation that complicates multi-pinch waveguides.

1) Downlink:

  • At direct-over-user placement, every user attains the same maximum downlink rate R0, independent of user index.The PA–user along-arm distance is dk = |ρs − xs,k|, and direct placement sets dk = 0.Source: 1) Downlink · p. 2

2) Uplink:

  • With exactly one PA on each arm, an uplink signal passes no other pinch on its way to the feed, so inter-antenna radiation is absent by construction.Residual scattering, inter-arm coupling, and switch leakage are neglected.Source: 2) Uplink · pp. 2–3

III. PIPELINED ANTENNA PRE-POSITIONING

PAP schedules service in frames and lets idle arms pre-position antennas while another arm transmits. The protocol uses cyclic arm selection and reverses each arm’s sweep at frame boundaries.

III. PIPELINED ANTENNA PRE-POSITIONING

  • Each frame serves every user once; the switch visits arms cyclically and skips arms whose users are already served.The batch size sets whether the schedule is interleaved or serves an arm’s users in one block.Source: §III · p. 3
  • At initialization, each arm sorts users by position and places its antenna at the first target; this setup is amortized over frames.Within a frame, an active arm serves users in sweep order, placing its antenna at each user’s target.Source: §III · p. 3
  • Idle arms move their antennas to their next targets while another arm is served, so re-pinching overlaps service on other arms and electronic guards.A move that finishes by the next eligible start adds no time; frame duration includes a closing guard.Source: §III · p. 3
  • At each frame boundary, arms reverse their sweep direction, so each antenna starts the next frame at the end of its users’ span without an approach move.The frame turnaround also covers the downlink–uplink transition.Source: §III · p. 3

IV. PERFORMANCE ANALYSIS

The analysis measures scheduling efficiency as the payload share of a frame and models mechanical re-pinching time from distance. Under alternating sweeps, sorted service minimizes per-frame travel.

IV. PERFORMANCE ANALYSIS

  • Scheduling efficiency is the fraction of frame time carrying payload; when every user receives rate R0, effective sum rate is ηR0.The frame duration is defined by the scheduling recurrence, and latency affects time rather than signal quality.Source: §IV · pp. 3–4

A. Re-pinching Latency Model

  • For a positive travel distance, re-pinching time is modeled using actuation time T0 and travel speed v, with illustrative values spanning 7 ms per 1 mm to about 2–20 s per 1 m.T0 includes pinch release, acceleration, and settling; the model treats T0 and v parametrically.Source: A. Re-pinching Latency Model · p. 4
  • Under alternating sweeps, each arm travels exactly its user-span distance D_s per frame, and no service order uses less motion.A monotone sweep between the span’s extremes achieves this distance; starting from an interior position requires extra travel.Source: A. Re-pinching Latency Model · p. 4

B. Interleaved Schedule and the Sufficient Number of Arms

The interleaved schedule has a closed-form efficiency bound, reaching the switching limit once the idle time between services can absorb re-pinching. The required number of arms depends on slot and movement times.

B. Interleaved Schedule and the Sufficient Number of Arms

  • The interleaved schedule attains its efficiency bound, and no schedule respecting the stated sweep orders and assumptions exceeds it; for q > 1, the switching limit is reached iff S ≥ S⋆.For q = 1, no move is needed and the switching limit is always reached.Source: B. Interleaved Schedule and the… · p. 4
  • When S ≥ S⋆, the idle interval W(S) is at least Trec and efficiency reaches η = Tslot/(Tslot + Tsw).The idle interval is W(S) = (S−1)(Tslot+Tsw)+Tsw; the threshold is S⋆ = (Trec + Tslot)/(Tslot + Tsw).Source: B. Interleaved Schedule and the… · p. 4
  • Alternating sweeps count the avoided move between frames exactly; for q > 1, this raises efficiency by at most q/(q−1) over its large-q limit and the saving vanishes at the switching limit.The maximum start-to-start interval between consecutive services of a user is 2Tframe − S(Tslot + Tsw).Source: B. Interleaved Schedule and the… · p. 4
  • At Trec = 50 ms and Tsw ≈ 0, S⋆ is 51 arms for 1 ms slots and 6 arms for 10 ms slots.A 50 ms re-pinch moves the antenna about 2 cm under T0 = 5 ms and v = 0.5 m/s; meter-scale traverses take seconds and require comparable service periods to absorb with few arms.Source: B. Interleaved Schedule and the… · p. 4

C. Burst Schedule

The burst schedule serves users in arm-sized blocks and accounts for each within-arm move alongside payload and guard time. When nonzero moves last at least as long as a guard, the total motion times determine the move contribution.

C. Burst Schedule

  • With B ≥ max_s K_s, an arm serves all its users in one block before the switch moves on.This is the burst schedule’s batch-size condition.Source: C. Burst Schedule · pp. 4–5
  • The burst schedule’s frame efficiency accounts for payloads, guards, and within-arm re-pinching across all arms.Each arm sweeps once per frame; a within-arm transition lasts max{Tsw, Trec(ℓs,j)} because motion overlaps the guard after the preceding payload.Source: C. Burst Schedule · p. 5
  • If every nonzero move lasts at least Tsw, individual move times enter the efficiency denominator only through total motion times M_s; for balanced loads and equal moves, interleaving is at least as efficient.The denominator reduces to KTslot + [K − Σ_s(N_s − 1)]Tsw + Σ_s M_s.Source: C. Burst Schedule · p. 5

D. Rate Analysis Against Fixed Placement

For q > 1, PAP matches a fixed-placement baseline only when its rate ratio and re-pinch latency satisfy the threshold conditions. At fixed q, baseline ratio, and timing parameters, tolerable latency increases by the same amount with each added arm.

D. Rate Analysis Against Fixed Placement

  • For users directly below their assigned arms, PAP's effective sum rate is η(S)R0, so it matches the baseline iff η(S) ≥ c.Both schemes use the same downlink/uplink time allocation.Source: D. Rate Analysis Against Fixed Placement · p. 5
  • PAP at least matches a fixed-placement baseline of rate ratio c iff c ≤ Tslot/(Tslot + Tsw) and Trec ≤ T*rec.Strict outperformance requires both inequalities to be strict.Source: D. Rate Analysis Against Fixed Placement · p. 5
  • At fixed q, c, Tslot, and Tsw, the tolerable latency increases by the same amount per added arm.The threshold follows from requiring both terms in the scheduling-efficiency expression to be at least c.Source: D. Rate Analysis Against Fixed Placement · p. 5

1) Fixed-Placement NOMA:

The fixed-placement NOMA reference serves two users continuously from one fixed pinch, with a near-optimal placement over one user. For the specified parameters, PAP exceeds this reference only near the switching limit.

1) Fixed-Placement NOMA:

  • The NOMA reference uses superposition coding to serve both users continuously during downlink time, neglecting guard overhead.A scan over placements finds placing the pinch over one user nearly optimal for these parameters.Source: 1) Fixed-Placement NOMA · p. 5
  • With near-user power fraction α, equal rates satisfy (1 + αγ0)(1 + αγ∆) = 1 + γ∆, and the positive root determines cN.The NOMA rate ratio is cN = 2 log2(1 + α*γ0) / log2(1 + γ0).Source: 1) Fixed-Placement NOMA · p. 5
  • cN = 0.9966: the fixed-placement NOMA sum rate is 0.34% below R0, so PAP exceeds it only near the switching limit.Matching the reference requires Tsw ≲ 3.4 µs per 1 ms slot.Source: 1) Fixed-Placement NOMA · p. 5

2) Fixed-Placement Star TDMA:

Fixed-placement star TDMA keeps each arm's pinch at its center, while PAP re-pinches to each user's position. Numerical results validate the scheduling-efficiency laws and show gains over fixed placement, especially for outer-endpoint users and at the fifth percentile as the number of arms grows.

2) Fixed-Placement Star TDMA:

  • The fixed-placement baseline centers each pinch, maximizing expected sum rate for users uniform along the arm, and incurs only Tsw per slot.Its effective sum-rate ratio includes the electronic guard overhead.Source: 2) Fixed-Placement Star TDMA · p. 5
  • Once latency is absorbed, re-pinching gains 6.7% in expected sum rate and 17% for the outer-endpoint user over fixed-placement TDMA.The fixed-placement reference has ρplace = 0.937 and ϵedge = 0.852 for every S because each arm spans D.Source: 2) Fixed-Placement Star TDMA · p. 5

V. NUMERICAL RESULTS

  • Interleaved and burst simulations agree with their closed-form efficiencies to within 10^-12, and interleaving reaches the switching limit at S ≥ S⋆.With equal moves and fixed users per arm, burst efficiency does not depend on S and remains below interleaving over the plotted range.Source: §V · p. 6, Figure 3 · p. 6
  • PAP's fifth-percentile gain over fixed placement rises from 3.5% at six arms to 9% at ten, while its mean gain stays near 4%.For q = 4 and Trec = 5Tslot, PAP absorbs latency at S⋆ = 6.Source: §V · p. 6, Figure 4 · p. 6
  • With eight arms, mean scheduling efficiency over a downlink–uplink cycle is 0.0116 at Tslot = 10 ms and 0.923 at Tslot = 2 s.These averages use random drops, unequal arm loads, distance-based move times, and skip empty arms.Source: §V · p. 6
  • Meter-scale motion needs service periods of seconds; the 50 ms moves assumed for Fig. 4 apply to closely spaced users.The random-drop test includes unequal arm loads and move times.Source: §V · p. 6

VI. CONCLUSION

S-PASS pairs a star of waveguide arms sharing one RF chain with PAP, which pre-positions idle-arm antennas while another arm transmits. The conclusion reports that adding arms expands tolerable latency, while re-pinching benefits outer-endpoint users most.

VI. CONCLUSION

  • Under balanced loads and equal move times, scheduling efficiency rises with arm count until the switching limit; six arms suffice for a 50 ms re-pinch and 10 ms slots.With q = 2 and Tsw/Tslot = 10^-4, PAP at least matches the two-user NOMA reference up to 0.0067 slot durations with one arm and 7.05 with eight.Source: §VI · p. 6
  • Once latency is absorbed for users directly below the arms, PAP gains 6.7% in expected sum rate and 17% for the outer-endpoint user over fixed-placement TDMA.With users spread over a disk and a five-slot move time, the fifth-percentile gain reaches 9% at ten arms.Source: §VI · p. 6
  • Random user drops with distance-based move times indicate meter-scale motion needs service periods of seconds.The conclusion identifies measured actuator parameters and the terahertz regime as future work.Source: §VI · p. 6
Loading 2610.29700v1…