{"id":"6fb2e98a-4bcf-4cca-a63d-8a1b355e8a8e","arxiv_id":"2504.21571","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A hybrid design with a few battery-powered beacons lets battery-free IoT devices synchronize on a shared slot schedule, improving data aggregation completion times by up to tens of times in some scenarios.","lead":"FreeBeacon adds a small number of battery-powered beacon devices to networks of battery-free IoT devices, allowing the battery-free devices to discover a coordinator and communicate on fixed time slots. Tests on prototypes and in simulation show faster data aggregation than random-delay and linear-probing baselines, though gains over the closest prior hybrid design are modest when no device failures occur.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1's guaranteed-discovery proof ignores collisions; the backoff in Alg. 1 either preserves the colliding slot (Bernoulli Tdist) or breaks the fixed-phase assumption, so the central guarantee is unproven for the multi-device case.","rationale":"The reader's weakest assumption was the single-beacon coverage limitation, which the paper itself acknowledges in Section V. My concern is different: the formal guarantee in Theorem 1, which is the foundation of the whole approach, has a proof gap for the collision case that the protocol is explicitly designed to handle. The reader did flag 'the collision guarantee in beacon discovery is asserted without proof' in the rationale, so there is partial agreement, but it was not identified as the weakest assumption. This concern is load-bearing because if the discovery guarantee does not hold under collisions, the claimed 'efficiency and robustness by construction' and the subsequent slot synchronization, structured aggregation, and failure-resilience arguments all lose their theoretical basis. The paper deserves credit for a simple and elegant co-prime observation, real prototype experiments, and substantial simulations; the issue is not that the approach is wrong, but that the central guarantee is stated more strongly than what is proven. A conditional verdict remains appropriate, requiring either a proof that covers the collision backoff or an explicit weakening of the guarantee. The concrete test I propose would decisively show whether the unconditional claim can survive a simple two-device collision scenario.","tokens_in":22997,"tokens_out":10692,"duration_ms":117015,"concrete_test":"Simulate Algorithm 1 with Tdist=5, Tb=2, and two devices with identical charging times that initially wake in the same slot, forcing RANDOM(0,1) to return 0 on every collision. If the two devices remain on the same distribution slot and neither ever receives a beacon reply, Theorem 1's unconditional guarantee is contradicted. Additionally, run the same simulation with RANDOM(0,1) uniform in [0,1) and record discovery time over many runs; if discovery has no bound or fails in some runs, the paper must state the guarantee as probabilistic or restricted to collision-free scenarios.","verdict_should_be":"UNCHANGED","load_bearing_attack":"FreeBeacon's central claim is 'guaranteed' beacon discovery (Theorem 1) and the efficiency built on it. The proof in Appendix B shows only that a single device waking at a fixed residue modulo Tdist will eventually coincide with the beacon, because co-primality makes the beacon index equidistributed. The multi-device case is not addressed. Section III-A states that collided devices receive 'an extra delay of Tdist slots randomly' (Alg. 1, line 13: delay += Tdist × RANDOM(0,1)). If RANDOM(0,1) is a Bernoulli multiplier of Tdist, the device's residue modulo Tdist is unchanged; two devices that collide at the same slot keep colliding on every distribution cycle, so the deterministic 'guaranteed' statement is false. If RANDOM(0,1) is fractional, the device's phase is altered so no fixed residue exists for the beacon to hit, and the Weyl-equidistribution argument no longer applies. In neither reading does the proof establish eventual discovery under collisions. Since beacon discovery is the foundation for slot synchronization, device-to-device communication, data aggregation, and failure recovery, this missing case is load-bearing.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper proposes FreeBeacon, a hybrid architecture for battery-free IoT in which a small number of battery-powered beacon devices provide a fixed duty-cycle beacon while battery-free devices align to a distribution cycle of length Tdist. The key theoretical claim (Theorem 1) is that if the beacon period Tb and Tdist are co-prime, every device is guaranteed to eventually meet the beacon, by an application of Weyl's equidistribution theorem. Once devices have discovered the beacon and corrected their slots, they can execute line-, tree-, and ring-based data aggregation using scheduled device-to-device communication, with recovery after device failures. The evaluation combines a five-node Riotee testbed, a controlled MOSFET-based testbed, Python simulations, and OMNeT++ simulations, comparing FreeBeacon against Find, Flync-Find, and Pulsar across several energy traces and failure rates.","tokens_in":23186,"tokens_out":13527,"duration_ms":141633,"significance":"If the discovery guarantee held for the full multi-device system, FreeBeacon would be a valuable and well-motivated design: it uses a tiny number of battery-powered coordinators to convert unpredictable intermittent communication into deterministic slot schedules, supports structured aggregation patterns, and handles device resets gracefully. The paper's strengths include a broad multi-platform evaluation, clear protocol pseudocode, and a formal single-device argument that correctly applies Weyl equidistribution when Tb and Tdist are co-prime. However, the flagship guarantee is not established for the multi-device case with collisions, and the abstract's 'consistently achieve an order of magnitude' claim for data aggregation is contradicted by the paper's own testbed numbers and by several near-parity results against Pulsar. The hybrid architecture and scheduling idea are promising, but the central claims need repair before the paper can be accepted.","major_comments":[{"comment":"Theorem 1 as stated ('If Tb and Tdist are co-prime, it is guaranteed that every device will eventually meet the beacon') is not proven for the multi-device case. The proof in Appendix B considers a single device with a fixed wake-up residue modulo Tdist and does not address collisions. Section III-A handles collisions with a backoff, but the pseudocode in Algorithm 1 line 13, 'delay <- delay + Tdist * RANDOM(0, 1)', is inconsistent with the prose 'an extra delay of Tdist slots randomly.' If RANDOM(0,1) returns 0 or 1, one natural reading, the colliding devices keep exactly the same residue modulo Tdist and collide again on every distribution cycle, so they may never discover the beacon. If RANDOM(0,1) returns a real in [0,1), the device wakes at non-integer offsets and no fixed residue modulo Tdist exists, so the Weyl equidistribution argument does not apply; if the delay is rounded to integer slots, the residue changes in an uncontrolled way. In neither reading does the paper establish 'guaranteed discovery regardless of possible collisions.' Because beacon discovery is the foundation for slot synchronization, device-to-device communication, aggregation, and failure recovery, this missing case is load-bearing. The authors should either modify the backoff to deterministically or almost surely separate collided devices into distinct residues and prove the corresponding guarantee, or explicitly restate the theorem as a single-device guarantee.","section":"Section III-A, Algorithm 1 lines 12-13, Appendix B"},{"comment":"The abstract and conclusions claim that FreeBeacon can 'consistently achieve an order of magnitude data aggregation efficiency' compared with state-of-the-art approaches. The data in Section IV-A do not support 'consistently.' On the Riotee testbed the reductions over Find are 54.73%, 23.25%, and 58.82%, which are factors of roughly 2.2, 1.3, and 2.4, not an order of magnitude. Against Pulsar in the large-scale line-aggregation results of Table I, FreeBeacon is nearly identical in several cases (e.g., 5,230 vs. 5,175 s for 6 devices in range [30,120], and 1,884 vs. 1,830 s for 30 devices in the same range). Many Find entries in Table I are incomplete, so the speedup ratios cited for those scenarios are not defined. The 'up to 29.5x' statement in the contributions refers to pairwise communication, not data aggregation. The wording should be narrowed to 'up to an order of magnitude in specific scenarios' unless the evaluation is expanded to substantiate a consistent aggregation-speedup claim.","section":"Abstract, Section IV-A (Fig. 8, Table I)"},{"comment":"The default configuration Tdist = 51 is selected as the best of three manually chosen values (30, 51, 100) on the same energy traces that are later used for the failure-rate evaluation, and Section V states that distribution-cycle adaptation is not supported. The paper therefore does not demonstrate that the reported gains are achievable without dataset-specific tuning, and it gives no deployment-time guidance for setting Tdist from the number of devices and the charging-time statistics. A sensitivity analysis over a wider parameter grid and an out-of-sample validation (selecting Tdist on one trace and testing on another) would be needed to support the claimed consistency across scenarios.","section":"Section IV-C-2, IV-D-2, Section V"}],"minor_comments":[{"comment":"The pseudocode has a typo: 'while ture do' should read 'while true do.'","section":"Algorithm 2, line 11"},{"comment":"The text says 'line, right, and tree' where 'ring' is intended; Figure 8 shows line, ring, and tree patterns.","section":"Section IV-A"},{"comment":"The caption reads 'Riotee-base' and should be 'Riotee-based.'","section":"Figure 9 caption"},{"comment":"The sentence 'all devices are guaranteed to be discovered within Tdist rounds' should define what a 'round' is; a full cycle of the beacon index sequence spans Tb * Tdist slots, not Tdist slots, and the distinction affects the expected discovery time.","section":"Section III-A"},{"comment":"The device logic uses Ti without making explicit that Ti is the current charging-cycle length; since the paper emphasizes that charging times vary per cycle, please state that line 11 is evaluated with the current cycle's charging time and explain how the device measures Ti in the absence of a synchronized clock before discovery.","section":"Algorithm 1"},{"comment":"The paper reports the default Tdist = 51 but does not state the corresponding Tb used in that experiment; please report the exact beacon cycle so the experiments are reproducible.","section":"Section IV-D-2"}],"recommendation":"major_revision","confidential_remarks":"The single-device equidistribution argument is correct, but the multi-device collision case is not a mere presentation gap: under the literal Algorithm 1 backoff, the deterministic guaranteed-discovery claim is either unproven or false, depending on the interpretation of RANDOM(0,1). I believe this is repairable, for example by introducing a backoff that re-selects a residue uniformly modulo Tdist and proving an almost-sure discovery guarantee, but the protocol and the theorem would both need to change. The abstract's 'order of magnitude' claim for data aggregation should also be corrected regardless of the technical repair. I recommend major revision rather than rejection because the hybrid architecture, the scheduling of line/tree/ring aggregation, and the breadth of the evaluation are useful contributions if the guarantee and the claims are made accurate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. FreeBeacon's core idea—a battery-powered beacon running a fixed cycle coprime to the battery-free devices' distribution cycle, so devices can synchronize to allocated slots—is genuinely new relative to Pulsar, and the aggregation patterns built on top are a useful step for battery-free IoT. But the paper's central guarantee is not proven as stated. The stress test lands: Theorem 1's proof covers only the collision-free single-device case, while the protocol promises discovery \"regardless of possible collisions.\" The backoff in Algorithm 1 either keeps a colliding device on the same residue modulo Tdist (if the random multiplier is 0 or 1) or destroys the fixed-phase assumption that the Weyl equidistribution argument needs. Either way, the multi-device collision case is unresolved, and since discovery underpins slot synchronization, communication, and recovery, this is a load-bearing gap.\n\nWhat deserves credit: the single-device argument is a correct application of equidistribution when Tb and Tdist are coprime; the line/tree/ring aggregation schedules are thoughtfully designed, with the tree collision-avoidance handled neatly; failure recovery by re-running discovery is a real practical benefit; and the evaluation is broad—Riotee prototype, controlled testbed, OMNeT++ simulations, Python simulations, and several real charging traces. The citation pattern is also fine: Pulsar, the closest prior work and from the same group, is directly compared rather than cited around.\n\nSoft spots, in proportion. The abstract's \"consistently achieve an order of magnitude\" claim is not supported by Table I's no-failure line/tree comparisons against Pulsar; those are mostly within a few percent. The default Tdist=51 is selected by performance on the same evaluation traces, which is a tuning-on-test-data concern. Table I has no error bars or variance information. Section V honestly admits the single-beacon, single-hop limitation and the open challenge of cross-beacon communication; that bounds the contribution but is not hidden. There is also an internal numerical mismatch: Section IV claims a 6.3x improvement for six devices in [30,120], but Table I shows FreeBeacon and Pulsar nearly tied on line aggregation there.\n\nWho it is for: researchers in battery-free IoT networking who want a concrete alternative to random-delay and linear-probing discovery. It deserves a serious referee. I would send it to review, not desk reject, but with a request for major revision: prove or repair the collision guarantee, temper the headline, add statistical detail, and stop tuning the default parameter on the test set.","headline":"Worth engaging: FreeBeacon's co-prime beacon discovery and aggregation schedules are a real step past Pulsar, but the collision-resilience guarantee is unproven and the headline claims outrun the data.","tokens_in":23772,"tokens_out":4067,"would_cite":true,"duration_ms":42506,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"FreeBeacon claims that a few battery-powered beacons with co-prime wake-up cycles make battery-free device discovery guaranteed, turning random intermittent encounters into scheduled, failure-resilient data aggregation.","keywords":["battery-free IoT","intermittently powered devices","beacon discovery","data aggregation","co-prime cycles","Weyl sequence","slot synchronization","energy harvesting"],"falsifier":"Run the paper's own large-scale scenario at a 5% device failure rate, for example 100 battery-free devices performing line aggregation with charging times in [30,120] slots and 1800 data items; if FreeBeacon cannot finish within the 300,000 second limit that the paper uses, the claimed failure resilience under realistic conditions is falsified.","tokens_in":1610,"feed_emoji":"📡","tokens_out":2761,"duration_ms":128268,"temperature":0.7,"pith_summary":"FreeBeacon's central claim is that the crippling uncertainty of battery-free IoT communication can be removed by adding a small number of battery-powered beacons, and that once a beacon is reachable, device discovery and structured data aggregation become deterministic. The mathematical core is that if the beacon's fixed cycle length $T_b$ is co-prime with the devices' shared distribution cycle $T_{dist}$, then the sequence of beacon-active slots is a Weyl sequence and is equidistributed modulo $T_{dist}$, so every device eventually meets the beacon. After that first contact, the device aligns to a unique slot on the shared cycle, and communication reduces to a sender jumping to the receiver's slot, which lets FreeBeacon implement line, tree, and ring data aggregation with recovery after power failures. A sympathetic reader would care because, if true, the protocol replaces random-guess neighbor discovery with a small, cheap source of certainty and delivers data-aggregation completion-time reductions of up to tens of times over existing approaches.","feed_headline":"One beacon ends the guessing game in battery-free IoT","feed_subtitle":"FreeBeacon adds a few battery beacons so intermittent devices meet, align slots, and aggregate data up to 29.5x faster.","key_machinery":"The central object is the pair of co-prime cycles: the beacon's fixed cycle $T_b$ and the shared distribution cycle $T_{dist}$ with length at least the number of devices. The beacon index is the remainder of multiples of $T_b$ modulo $T_{dist}$, which forms a Weyl sequence and therefore visits every slot of the distribution cycle with equal frequency when $T_b$ and $T_{dist}$ are co-prime. This property guarantees eventual device-beacon overlap, gives every device a unique slot on the shared cycle, and converts device-to-device communication into a simple slot jump with rollback.","core_discovery":"FreeBeacon's central claim is that injecting a small amount of certainty, in the form of one always-on battery-powered beacon with a fixed duty cycle, transforms battery-free neighbor discovery from a random-guessing problem into a deterministic scheduling problem. The enabling result is Theorem 1: when the beacon cycle length $T_b$ and the distribution cycle length $T_{dist}$ are co-prime, the beacon index follows a Weyl sequence that is equidistributed modulo $T_{dist}$, so every battery-free device that wakes according to the distribution cycle is guaranteed to eventually coincide with the beacon. Once a device hears the beacon, it learns the beacon's current slot index and corrects its wake phase so that it occupies a unique pre-assigned slot on the shared distribution cycle; communication then becomes a sender jumping to the receiver's slot, exchanging data, and rolling back. FreeBeacon uses this slot machinery to implement line, tree, and ring data aggregation, and the evaluation reports that this removes rediscovery overhead after power failures, yielding up to 29.5x lower completion time than random-guess baselines.","pith_inferences":["Beyond the paper, the same coprime-period guarantee should apply to any pair of intermittently active components that share a periodic schedule, so the mechanism could be reused for coordinated sensing or actuation in batteryless systems, not only for communication.","The one-slot-per-device allocation implies that the distribution cycle must grow at least linearly with the number of devices, so large networks face an inherent trade-off between slot count and per-round aggregation latency; the paper's evaluations stop at 100 devices and do not quantify this scaling limit.","A testable extension would be an adaptive distribution cycle: when the beacon detects new or departed devices, it could broadcast a new $T_{dist}$ and have devices recompute their offsets, an operation the paper leaves for future work.","Because the beacon only needs to receive discovery messages and broadcast slot indices, FreeBeacon could be layered over existing low-power physical layers such as BLE or backscatter without changing the scheduling logic; the paper's testbeds only exercise BLE-style radios."],"forward_implications":["With one reachable beacon, every battery-free device is guaranteed to discover it eventually, and after a full reset the device recovers synchronization by rerunning the same discovery protocol.","Device-to-device communication becomes a slot jump: the sender extends its charging time until the receiver's slot, exchanges the message, and rolls back, so no random discovery delays are involved.","Line, tree, and ring data aggregation all run on the shared slot schedule, and collisions are avoided by construction because each device holds a unique slot in each round.","Failure resilience is automatic for senders, because the beacon sniffs the channel and broadcasts the current slot index whenever a sender transmits from a wrong slot, and receivers periodically query the beacon to correct their own slots.","Selecting the smallest integer coprime to $T_{dist}$ as the beacon cycle $T_b$ minimizes synchronization time, giving a concrete deployment rule."],"supporting_citations":[{"why":"Supplies the Find and Flync-Find random-delay baselines whose parameter-tuning fragility and poor structured-communication performance motivate FreeBeacon.","marker":"[11]"},{"why":"Supplies the Weyl-sequence equidistribution theorem that Theorem 1 relies on for guaranteed beacon discovery.","marker":"[22]"},{"why":"Provides the real-world charging time traces from cars, office, stairs, and washer used to evaluate FreeBeacon under realistic energy variability.","marker":"[23]"},{"why":"Pulsar is the closest hybrid-design baseline, using linear probing for beacon discovery, against which FreeBeacon's number-theoretic discovery is compared.","marker":"[25]"},{"why":"The Riotee platform is the hardware used in FreeBeacon's real-world five-device testbed for data aggregation experiments.","marker":"[26]"}],"fun_headline_variants":["One beacon turns IoT guessing into deterministic slots","FreeBeacon: deterministic scheduling for battery-free IoT","Battery-free IoT stops guessing: beacons impose order","Co-prime beacons make battery-free aggregation 29.5x faster","A single beacon schedules battery-free IoT data aggregation"],"cache_read_input_tokens":25856,"weakest_assumption_plain":"The design assumes every battery-free device can directly reach the beacon and its intended peer by radio, with no multi-hop or coverage model, so a device that cannot hear the beacon falls outside the discovery and slot-correction guarantees.","fun_headline_variants_meta":{"raw":{"variants":["One beacon turns IoT guessing into deterministic slots","FreeBeacon: deterministic scheduling for battery-free IoT","Battery-free IoT stops guessing: beacons impose order","Co-prime beacons make battery-free aggregation 29.5x faster","A single beacon schedules battery-free IoT data aggregation"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000412,"raw_usage":{"total_tokens":2140,"prompt_tokens":964,"completion_tokens":1176,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":580,"completion_tokens_details":{"reasoning_tokens":1107}},"tokens_in":580,"tokens_out":1176,"duration_ms":9116,"temperature":1.0,"reasoning_tokens":1107,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T05:00:17.003877+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the paper's own large-scale scenario at a 5% device failure rate, for example 100 battery-free devices performing line aggregation with charging times in [30,120] slots and 1800 data items; if FreeBeacon cannot finish within the 300,000 second limit that the paper uses, the claimed failure resilience under realistic conditions is falsified.","supporting_citations":[{"cited_title":"Bootstrapping battery -free wireless networks: Efficient neighbor discovery and synchronization in the face of intermittency,","cited_arxiv_id":null,"evidence_quote":"Supplies the Find and Flync-Find random-delay baselines whose parameter-tuning fragility and poor structured-communication performance motivate FreeBeacon."},{"cited_title":"U¨ ber die gleichverteilung von zahlen mod. eins,","cited_arxiv_id":null,"evidence_quote":"Supplies the Weyl-sequence equidistribution theorem that Theorem 1 relies on for guaranteed beacon discovery."},{"cited_title":"Demo abstract: Building battery -free devices with riotee,","cited_arxiv_id":null,"evidence_quote":"The Riotee platform is the hardware used in FreeBeacon's real-world five-device testbed for data aggregation experiments."}],"review_version":1}