SmartBAN on Silicon by Structured Behavioral Modeling

PREPRINT

1

SmartBAN on Silicon by Structured Behavioral Modeling

arXiv:2608.29036v1 [cs.NI] 29 Aug 2026

Masato Yoshimi , Student Member, IEEE, Takahiro Ito , Member, IEEE, Kento Tanaka , and Hirokazu Tanaka , Senior Member, IEEE

To address these requirements, SmartBAN [3]–[6] was standardized as a lightweight WBAN protocol whose timedivision multiple access (TDMA)-based channel access [7] guarantees each sensor node a deterministic uplink slot. While Bluetooth Low Energy (BLE) [8], [9] dominates commercial wearables, its connection-event scheduling cannot maintain deterministic timing as the number of peripherals grows [10]. Despite mature specifications covering physical layer (PHY) and media access control (MAC) layers, no WBAN standard— neither SmartBAN nor IEEE 802.15.6 [11]—has been fully realized on a commercial off-the-shelf wireless system-onchip (SoC); prior work is reviewed in Section II. WBAN standards define frame formats and message sequences but leave internal device behavior unspecified, creating a specification-toimplementation gap that may have limited SmartBAN adoption relative to BLE. We bridge this gap by modeling the unspecified behaviors as finite automata, realizing the model on a commercial wireless SoC, and validating the system through multi-axis experiments.1 The main contributions are: 1) We describe four specification gaps and address them as two Mealy-type finite automata—Hub (MH ) and Node (MN )— whose transition tables map directly to firmware dispatch logic: phase control and connection lifecycle are captured at the model level, while slotlevel timing and scheduling policy are resolved through structured realization. Index Terms—Body area network, SmartBAN, protocol implementation, Internet of Medical Things (IoMT), wireless sensor 2) To our knowledge, the first operational SmartBAN network. realization on a commercial wireless SoC, translating the behavioral model into a four-layer software architecture with a hardware-independent portability boundary. I. I NTRODUCTION 3) Multi-axis experimental validation with sixteen concurrent sensor nodes—behavioral conformance, standardONTINUOUS physiological monitoring through bodycompliant communication quality, and long-term worn sensors underpins the Internet of Medical Things stability—plus a same-SoC BLE comparison that quan(IoMT) [1], [2]. As applications span clinical diagnostics and tifies the determinism–efficiency tradeoff, covering the everyday wellness, wireless body area networks (WBANs) Node-to-Hub connection and uplink data paths required must deliver low latency, low power, multi-node support, and for sensor data collection (cf. Section II). reliability under body shadowing. Section II reviews the standard and identifies four specificaThis work has been submitted to the IEEE for possible publication. Copyright tion gaps. Sections III–IV present the behavioral model and may be transferred without notice, after which this version may no longer be its realization; Sections V–VI report and discuss experimental accessible. (Corresponding author: Masato Yoshimi.) M. Yoshimi, T. Ito, and H. Tanaka are with Graduate School of In- results. Abstract—Wireless body area networks (WBANs) are a key enabling technology for the Internet of Medical Things (IoMT). SmartBAN, standardized by ETSI and later adopted as an IEC international standard, defines a lightweight WBAN protocol with time-division multiple access (TDMA)-based physical (PHY) and media access control (MAC) layers, yet no implementation on commercial hardware has been reported. The standard specifies frame formats and channel structure but leaves internal device behaviors unspecified: phase control and connection lifecycle lack transition logic, while slot-level timing and scheduling policy lack parametric guidance. This paper addresses these omissions through structured behavioral modeling and modeldriven implementation. Two Mealy-type finite automata—one for the Hub (3 states, 5 transitions), one for each Node (5 states, 8 transitions)—capture phase control and connection lifecycle as a hardware-independent design blueprint whose transition tables map directly to firmware dispatch logic; slot-level timing and scheduling policy are resolved through realization on the nRF54L15, a commercial Arm Cortex-M33 wireless system-onchip (SoC) running Zephyr real-time operating system (RTOS). Experiments with sixteen concurrently scheduled sensor nodes over 25 hours validate the design for the initial connection and uplink data paths: all 13 modeled transitions were exercised with sub-millisecond per-slot timing jitter (P99 < 754 µs, slotindependent across all 16 slots), 99.99% packet delivery, and autonomous disconnection recovery. A same-SoC Bluetooth Low Energy (BLE) comparison quantifies the determinism–efficiency tradeoff: SmartBAN achieves substantially lower timing jitter at higher energy cost, the majority of which is attributable to software radio processing rather than the protocol-level duty cycle.

C

formation Sciences, Hiroshima City University, Hiroshima, Japan (e-mail: myoshimi@e.hiroshima-cu.ac.jp). M. Yoshimi and K. Tanaka are with Strategic Technology Center, Technology SBU, TISI Inc., Tokyo, Japan. H. Tanaka is also with Graduate School of Informatics, Osaka Metropolitan University, Osaka, Japan.

1 Earlier implementation-driven prototypes targeted sensor applications— RSSI-based posture sensing on TI CC2650 [12] and wearable IMU streaming with multiple nodes [13]. This paper instead develops the protocol methodology: structured behavioral modeling, model-driven realization, 16-node long-term validation, and same-SoC BLE comparison.

PREPRINT

2

TABLE I S IX P RINCIPAL MAC O PERATIONS OF S MART BAN (C LAUSES PER ETSI TS 103 325 [3]) ID

Operation

(1) (2a) (2b) (3) (4) (5) (6)

Initial connection Uplink (Scheduled Access Period) Uplink (Contention Management Period) Supplementary downlink Slot re-assignment Data channel migration Disconnection

Clause

Coverage

7.2.2 7.3.1 7.3.2 7.4 7.5 7.6 6.2.6, 6.2.7

This paper This paper This paper Extension Extension Extension Extension

TABLE II C OMPARISON OF WBAN AND S MART BAN S TUDIES Ref.

Protocol

Approach

Platform

Valid.

[16]

802.15.6

[14] [15] [21], [22] [23]–[25] [26] [27] [28]

802.15.6 802.15.6 SmartBAN SmartBAN SmartBAN SmartBAN SmartBAN

MAC+PHY SoC Research ASIC 400 MHz, (40 nm CMOS) (IMEC/Fujitsu) not OTS NB-PHY design FPGA PHY only PHY validation SDR PHY only Coexistence OMNeT++ Sim., ≤10 N MAC optimization Simulation Sim. only Throughput model Analytical — Battery optim. Simulation Sim. only BLE comparison Simulation Sim. only

This work SmartBAN Behavioral model Commercial 16-node, II. S MART BAN S PECIFICATIONS AND D ESIGN G APS → impl. SoC 25 h HW A. Protocol Overview SmartBAN is a WBAN protocol standardized by ETSI [3], [4] and later adopted as IEC 63203-801 [5], [6]. It defines a (connected, unconnected) are named, but the complete star-topology network in which a single Hub coordinates up transition logic—triggers, guards, timeout-based disconto 16 sensor Nodes [7]. nection detection, and autonomous recovery procedures— The PHY [4] defines 40 channels at 2 MHz spacing in is absent from the specifications. the 2.4 GHz ISM band (2402–2480 MHz), using Gaussian 4) Scheduling policy. Although the Hub assigns slots frequency-shift keying (GFSK) modulation at 1 Mbit s−1 via C-Assignment, the standard does not prescribe the with optional Bose–Chaudhuri–Hocquenghem (BCH) forward allocation algorithm—fairness criteria, priority handling, error correction. Three of these channels serve as control or dynamic reallocation upon node join/leave events. channels (CCH) for frequency-diverse beacon advertisement; These gaps concern behaviors required for minimal singlea single data channel (DCH) carries scheduled traffic. channel star-topology operation; optional features (relay, coexThe MAC [3] structures communication into a repeating istence, key management) lie outside this scope. The four gaps inter-beacon interval (IBI). Each IBI comprises a D-Beacon differ in nature: Gaps 1 and 3 involve multi-state coordination carrying a timestamp and slot-allocation map, a scheduled logic that can be captured independently of any hardware target, access phase (SAP) of deterministic TDMA uplink slots whereas Gaps 2 and 4 require parametric choices tied to a with immediate acknowledgment (I-Ack), and a contention concrete platform and scheduling policy. management period (CMP) where unconnected Nodes send connection requests via slotted Aloha. This structure guarantees each connected Node a deterministic uplink opportunity once C. Related Work As Table II shows, no WBAN standard has been per IBI. Viewing the MAC specification [3] from a behavioral fully realized on a commercial off-the-shelf SoC [1], [2]; perspective yields six principal operations summarized in IEEE 802.15.6 [11] realizations remain limited to partial PHYTable I; this paper realizes the three required for Node-to-Hub layer prototypes on FPGA or SDR platforms [14], [15], or to connection and uplink data collection—(1), (2a), and (2b)— research-grade multi-mode radio SoCs targeting the 400 MHz while operations (3)–(6) are accommodated as bounded exten- MICS band [16], and BLE [8], [9] dominates commercial wearables through connection-event scheduling rather than sions of the behavioral model (Section III). TDMA [10], [17]. Formal methods have been applied to WBAN protocol verification (timed automata, statistical model B. Specification Gaps While the standard [3], [4] defines frame formats, channel checking [18], [19]), but not to resolve specification gaps and parameters, and the IBI structure, it leaves four categories of inform implementation. All SmartBAN studies to date rely on simulation or analytical modeling; Hämäläinen et al. [20] internal device behavior unspecified: confirm continued standardization interest but no hardware 1) Phase control. The standard distinguishes CCH for realization. beacon advertisement and DCH for scheduled communication, but the internal mechanism for switching between III. B EHAVIORAL M ODELING them—CCH entry, network re-acquisition after beacon A. Approach and Methodology loss, and DCH resumption upon C-Beacon reception— remains unspecified. This section describes the hardware-independent behavior 2) Slot-level timing. The standard defines slot boundaries of SmartBAN Hub and Node devices as a structured model and IBI duration yet provides no guidance on how Nodes that addresses the phase-control and connection-lifecycle gaps derive per-slot radio timing from D-Beacon timestamps (Section II-B); slot-level timing and scheduling policy, which or compensate for inter-IBI clock drift. depend on platform choices, are deferred to Section IV. Rather 3) Connection lifecycle. Connection-related frame types than a formal specification intended for model checking, the (C-Request, C-Assignment, I-Ack) and endpoint states model serves as a design blueprint whose transition-output

PREPRINT

semantics map one-to-one to event-driven firmware dispatch branches, making model and implementation directly coderivable. While timed automata have been applied to WBAN protocol verification [18], [19], we adopt Mealy-type finite automata [29], [30] optimized for implementability: each transition produces an output action, and the resulting tables translate directly to firmware dispatch logic with no intermediate verification step. Timed automata would enable exhaustive temporal verification but introduce specification complexity (clock variables, invariants) beyond what a design blueprint requires; timing parameters are instead validated empirically in Section V. We adopt compact notation where T maps (state, event) pairs to (action, next-state) pairs; a single super-state exit (τN 8 ) follows Harel Statechart notation [31], [32] for brevity. The model abstracts MAC-layer details such as BCH coding, counter thresholds, and scheduling policy, treating them as atomic side-effects within individual transitions; this separation preserves behavioral correctness while keeping the state space tractable. Each device model captures the logic governing channel-phase alternation (CCH and DCH) and sub-state transitions in response to received frames and timing events. B. Behavioral Model of SmartBAN Devices

3

TABLE III H UB AUTOMATON T RANSITION RULES Label State SH

Event EH

Action AH

Next State SH

τH1 τH2 τH3 τH4 τH5

Inactive_End CM P _End RX_C_Req Slot_End RX_U plink

T X_D_Beacon T X_C_Beacon T X_I_Ack T X_C_Assign T X_I_Ack

(DCH, IDLE) CCH (DCH, ASGD) (DCH, IDLE) (DCH, IDLE)

CCH (DCH, IDLE) (DCH, IDLE) (DCH, ASGD) (DCH, IDLE)

TABLE IV N ODE AUTOMATON T RANSITION RULES Label State SN

Event EN

Action AN

Next State SN

τN 1 τN 2 τN 3 * τN 4 τN 5 τN 6 τN 7 τN 8 †

RX_C_Beacon RX_D_Beacon RX_I_Ack RX_C_Assign RX_D_Beacon RX_D_Beacon RX_I_Ack T imeout

T X_C_Req T X_U plink T X_U plink -

(DCH, IDLE) (DCH, REQ) (DCH, REQ) (DCH, CONN) (DCH, UPLINK) (DCH, UPLINK) (DCH, CONN) CCH

CCH (DCH, IDLE) (DCH, REQ) (DCH, REQ) (DCH, CONN) (DCH, UPLINK) (DCH, UPLINK) SN \ {CCH}

* An I-Ack received while in REQ (before C-Assign arrives) is explicitly

consumed rather than left undefined, ensuring complete input coverage for the REQ state; the Node continues to await the management-plane C-Assignment response. †τ N 8 is a Harel super-state exit from the DCH region. The current SmartBAN specification defines no explicit disconnect command; connection termination relies on beacon-loss timeout.

The SmartBAN protocol behavior is described using two finite automata: one for the Hub device and one for the Node devices. Each is specified as a 5-tuple (S, E, A, s0 , T ), where (DCH, U P LIN K). There the Node retransmits on each S is the state set, E is the event set, A is the action set, s0 ∈ S subsequent D-Beacon (τN 6 ) until I-Ack (τN 7 ) returns it to is the initial state, and T is the transition function. CONN for the next uplink. The Hub coordinates medium access by alternating between Definition 2 (Node Automaton): The Node automaton Control Channel (CCH) and Data Channel (DCH) phases. MN = (SN , EN , AN , s0,N , TN ) is defined as: Definition 1 (Hub Automaton): The Hub automaton MH = 1) State set: SN = {CCH} ∪ ({DCH} × (SH , EH , AH , s0,H , TH ) is defined as: {IDLE, REQ, CON N, U P LIN K}), where CCH 1) State set: abbreviates (CCH, −) as in Definition 1, giving five SH = {CCH} ∪ ({DCH} × {IDLE, ASGD}), states of uniform product type. where CCH abbreviates the pair (CCH, −) with an 2) Event set: irrelevant sub-state, giving three states of uniform EN = {T imeout, RX_C_Beacon, RX_D_Beacon, product type. RX_I_Ack, RX_C_Assign}. 2) Event set: 3) Action set: AN = {T X_U plink, T X_C_Req, −}, EH = {Inactive_End, Slot_End, CM P _End, where “−” denotes no output action. RX_U plink, RX_C_Req}. 4) Initial state: s0,N = CCH. 3) Action set: AH = {T X_C_Beacon, T X_D_Beacon, 5) Transition function: TN : SN × EN ⇀ AN × SN is a T X_I_Ack, T X_C_Assign}. partial function (Table IV); undefined pairs are silently 4) Initial state: s0,H = (DCH, IDLE). The first ignored. CM P _End event triggers τH2 to transition to CCH. Both state machines are single-valued (deterministic in the 5) Transition function: TH : SH × EH ⇀ AH × SH is a automata-theoretic sense): for every (s, e) pair in the domain partial function (Table III); undefined pairs are silently of TH (respectively TN ), Tables III–IV specify exactly one ignored. (a, s′ ) outcome (τN 8 expands to four flat transitions per the The Node discovers the Hub via CCH, acquires tim- super-state exit semantics, each single-valued); events outside ing from D-Beacons, and negotiates a connection through the domain produce no state change. Fig. 1 visualizes both the C-Request/C-Assignment exchange. Upon receiving a models as Harel Statecharts, encapsulating DCH sub-states into C-Beacon (τN 1 ), the Node enters (DCH, IDLE)—timing super-states. Auxiliary arrows show how one device’s output acquired without an active connection—and issues a connection actions trigger the other’s input events. request on the next D-Beacon (τN 2 ). Once C-Assignment The Hub and Node models interact via broadcast coupling is received (τN 4 ), the Node enters (DCH, CON N ) and over a shared wireless channel: each transmit action maps transmits its first uplink on the next D-Beacon (τN 5 ), entering to the corresponding receive event in the other device (e.g.,

PREPRINT

4

Hub (

)

Application Layer

DCH CCH

Sensor acquisition, data forwarding IDLE

ASGD

MAC Layer D-Beacon Uplink

C-Beacon

Node (

CCH

)

Behavioral automata (SMF), IBI engine

C-Req I-Ack

I-Ack

Hardwareindependent

C-Assign

PHY Layer

DCH

BCH codec, CRC

IDLE

Portabilityboundary

REQ

Radio HAL GFSK, DMA UPLINK

Timer HAL RTC

Peripheral HAL SPI, I2 C

Platformspecific

CONN

nRF54L15 SoC (Cortex-M33, 2.4 GHz Radio, GRTC)

Fig. 1. Behavioral model of SmartBAN devices: Visualization of Hub-Node state machine interactions based on Definitions 1 and 2.

Fig. 2. Four-layer software architecture deployed on the nRF54L15. The MAC and PHY layers are designed to be hardware-independent; porting to a different SoC requires replacing only the HAL modules below the portability boundary.

T X_C_Beacon ∈ AH triggers RX_C_Beacon ∈ EN ). The N Node replicas execute independently—address-based frame filtering ensures each processes only its own Hub frames— virtualizes register-level access and peripheral configuration as yielding a system comprising one Hub and N independent the portability boundary. (1) (N ) replicas (MH with MN , . . . , MN ). If a connection request arrives while the Hub is already processing an assignment B. Behavioral Model Realization (state (DCH, ASGD)), it falls outside the domain of TH Each device state machine of Section III is realized within and is silently ignored; the requesting Node will retry via its the MAC layer, with transitions driven by two event sources: exponential backoff mechanism (Section IV-D). timer-based slot boundaries and frame reception. Each transition Design properties. The model ensures progress by three row of the Hub and Node tables (Tables III–IV) maps to a mechanisms: (i) the Hub’s CM P _End/Inactive_End timer dispatch branch in the implementation, preserving one-to-one cycle fires unconditionally, ensuring the Hub always progresses structural correspondence that enables behavioral conformance between CCH and DCH; (ii) the Node’s T imeout (τN 8 ) pro- verification. vides a universal escape from any DCH sub-state back to CCH, Timer-driven transitions. A real-time counter (RTC) comafter which the reconnection path τN 1 → τN 2 → τN 4 → τN 5 pare channel determines each slot boundary. On D-Beacon re-enters the CONN–UPLINK data-transfer cycle via (DCH, reception, the Node reprograms the RTC compare register UPLINK); (iii) the CONN–UPLINK retransmission cycle (τN 5 – from the beacon timestamp, compensating for inter-IBI clock τN 7 ) guarantees uplink acknowledgment: a Node retransmits drift and resolving the slot-level timing gap. A slot counter, the same request on each D-Beacon (τN 6 ) until I-Ack is incremented at every boundary, drives phase transitions (e.g., received (τN 7 ), then returns to CONN for the next data request. τH1 , τH2 , τN 8 ). At the DCH/CCH boundary, the radio is Because no inter-node state coupling exists, each of N replicas reconfigured from the data frequency to the control frequency. follows the same single-node behavior. Because a single radio transceiver handles both C-Beacon and Extensibility. Operations (3)–(6) of Table I each map to a D-Beacon traffic, the two frame types are inherently sepabounded extension of MH or MN —a small number of states, rated by the phase structure of the automaton—no additional events, and transitions tied to existing D-Beacon fields or arbitration is required. new management frames—preserving single-valuedness and Event-driven transitions. The PHY layer validates the bounded reachability. cyclic redundancy check (CRC) and performs BCH decoding on each received frame. For valid frames, the MAC layer dispatches to the appropriate handler based on the frame type IV. I MPLEMENTATION and subtype fields extracted from the Frame Control header. A. Software Architecture Each handler references the device’s current state to select the The hardware-dependent gaps (slot-level timing and schedul- appropriate transition (e.g., τH3 , τH5 for the Hub; τN 1 –τN 7 ing policy; Gaps 2 and 4) are resolved by realizing the for the Node). The I-Ack frame is pre-assembled before each behavioral model of Section III on a commercial wireless SAP slot to achieve sub-millisecond TX/RX turnaround for SoC through the four-layer software architecture of Fig. 2. acknowledged uplink reception. The Application layer handles sensor acquisition; the MAC layer manages connection state, slot scheduling, and frame assembly; the PHY layer handles radio configuration and forward error correction; a hardware abstraction layer (HAL)

C. Implementation Details PHY layer. Following the SmartBAN PHY specification [4], the radio operates in proprietary GFSK mode (h=0.5, BT=0.5,

PREPRINT

5

1 Mbit s−1 ) with a 32-bit preamble, bypassing the SoC’s BLE stack. BCH(127,113,t=2) forward error correction protects all frames, using pre-computed GF(27 ) lookup tables (896 B for exponentiation, logarithm, and parity tables). In this implementation, a single fixed data channel carries all scheduled traffic; during CCH, the Hub transmits C-Beacons on control channel 0, and the Node sequentially scans the three standarddefined control channels. MAC layer. Both Hub and Node maintain a Node Table— a per-connection context structure that tracks the connection sub-state defined in the transition tables—so that the set of connected Nodes is determined at runtime rather than fixed at compile time. All timing and configuration values (slot width, inter-frame spacing, IBI, etc.) are parameterized as compiletime constants. D. Scheduling and Timeout Policies

Fig. 3. Sensor node: XIAO nRF54L15 [33] (21 × 17.8 mm, ∼USD 6) with Nordic nRF54L15 [34] SoC (Arm Cortex-M33, 128 MHz, 1.5 MB Flash, 256 kB RAM) running Zephyr RTOS v4.2 [35] / nRF Connect SDK [36] v3.2.1, custom PCB carrying an LSM9DS1 IMU (I2 C), and battery holder. Firmware footprint: 59 kB/63 kB Flash (Hub/Node), ∼35 kB RAM each (4% / 14% of available capacity).

Slot allocation. The Hub uses first-fit sequential allocation from slot 1, assigning Nsap consecutive free slots to minimize allocation latency for up to Nmax = 16 nodes. B. Behavioral Conformance Timeout and recovery. Three timeout parameters govern Behavioral conformance is assessed by comparing observed connection robustness. On the Node side, a Connection Request state transitions against the structured behavioral model of is retried up to Ncreq times with exponential backoff (one Section III. The Hub firmware logs each transition (state, event, to eight IBIs); failure triggers τN 8 (Table IV). After I-Ack action, next state) via UART; a post-hoc script verifies each reception, the Node waits up to Ncass IBIs for C-Assignment; timeout triggers the same recovery. Once connected, the Node entry against Tables III–IV. All 13 transition rows fired at monitors D-Beacon reception: if Nmiss consecutive beacons are least once during the 25 h session, with per-transition counts missed, the Node transitions to CCH (τN 8 ) and autonomously spanning 1 (τN 8 ) to 6.55×106 (τH5 ); no undefined (state, event) rescans. On the Hub side, if a connected Node produces no pairs were observed. Connection lifecycle. One disconnection event triggered uplink for Nmiss IBIs, the Hub releases the allocated slots (an implementation-level policy outside the Hub automaton). τN 8 via the timeout mechanism of Section IV-D; the Node All timeout values are compile-time constants (Ncreq = 5, returned to CCH and completed the full reconnection sequence (τN 1 → τN 2 → τN 4 → τN 5 ) as modeled. Ncass = 3, Nmiss = 10). Uplink retransmission. The CONN↔UPLINK cycle (TaTiming precision. Fig. 4 shows the timing jitter distribution. ble IV) is realized as a single dispatch branch keyed on The measured inter-beacon interval was 220.005 ± 0.035 ms, the I-Ack reception flag, re-using the same TX buffer until a drift of 23 ppm.2 Per-slot P99 is slot-independent at 737– acknowledgment. V. E XPERIMENTAL E VALUATION Experiments target operations (1), (2a), and (2b) of Table I, exercising the 13 transitions of MH and MN .

2With N > 16,000 samples per slot, the 95% confidence interval for σ is narrower than ±3 µs.

TABLE V T IMING AND C ONFIGURATION PARAMETERS

A. Experimental Setup Fig. 3 shows the sensor node used in all experiments. The Hub runs identical hardware and firmware; it additionally connects to a host PC via UART for experiment logging. All experiments were performed indoors at 1 m Hub–Node distance in a shared laboratory with typical ambient Wi-Fi and Bluetooth activity. Table V summarizes the timing and configuration parameters common to all experiments. A Nordic PPK2 [37] in source mode supplied 3.8 V via the BAT+ pad at 100 kS/s (0.2 µA resolution, ±5% systematic accuracy) for board-level current measurement. Only the current-profiling firmware disables non-protocol peripherals to isolate the energy footprint (four GPIO channels mark protocol states for perIBI accounting); all other experiments use the full application firmware with the IMU sampled at 100 Hz.

Parameter

Symbol Value

Note

Standard constants Min. time unit Tmin Slot length Lslot Max nodes Nmax

625 µs 16 16

Fixed One of six standard options Fixed

Design parameters Inter-frame space TIFS Slots per node Nsap CMP slots Ncmp Inactive slots Ninact

150 µs 1 2 3

Fixed (TIFS) Configurable Configurable Configurable

Derived values Slot duration SAP slots Slots per IBI IBI duration

10 ms 16 22 220 ms

Lslot × Tmin Nmax × Nsap 1 + NS + Ncmp + Ninact LD × TS

TS NS LD TD

Probability density

PREPRINT

6

10

3

10

1

SmartBAN BLE N=1 BLE N=16

σ=0.37

10

−1

10

−3

10

−5

10

−7

×29 ×280

σ=104

σ=10.9

10

−3

10

−2

10

−1

10

0

10

1

10

2

10

3

Timing jitter (ms) Fig. 4. Timing jitter distribution for SmartBAN (slot 1) and BLE (N =1 and N =16), each measured over 1 h. Dashed vertical lines mark σ for each protocol. SmartBAN concentrates all samples within 0.8 ms, while BLE exhibits 29–280× higher jitter (σ). BLE comparison is discussed in Section VI-B.

754 µs; for slot 1 (Fig. 4), σ = 371 µs. In other slots, <0.1% of frames arrive ∼1 IBI late, inflating σ but not P99 . Frame exchange. Per-node packet delivery ratio (PDR) across the 16 Nodes is tight (coefficient of variation, CV = 0.004%), confirming equitable TDMA slot utilization.

D. Long-Term Stability The 25 h continuous test (BCH enabled, setup described in Section V-A) verified sustained operation without timedependent degradation (memory leaks, timer drift, buffer exhaustion). All Nodes transmitted live IMU sensor data. Aggregate PDR exceeded 99.99% (per-node range 99.983– 99.998%) with mean RSSI ranging from −62 to −42 dBm; per-node PDR showed no correlation with RSSI within this range. Only one disconnection occurred: Node 2 (mean RSSI −62 dBm, instantaneous minimum −105 dBm) was briefly disconnected and reconnected within 264 ms (a single IBI). The resulting maximum inter-frame gap of 2.4 s is consistent with timeout detection (Nmiss × TD ≈ 2.2 s) plus reconnection latency, suggesting that the disconnection was caused by transient signal attenuation from the surrounding environment rather than by a steady-state link-budget deficit. VI. D ISCUSSION A. Model Validation We assess the validity of the Ruan et al. analytical model against the measured current profile (Fig. 6, Table VI). Following Ruan et al. [38], the average current is evaluated using a three-state weighted average: (a) Iavg = ITX · dTX + IRX · dRX + Iidle · (1 − dTX − dRX ), (1)

where dTX and dRX are the TX and RX duty cycles (Table VI: dTX = 1.02%, dRX = 6.79%). Using the nominal values C. Standard-Compliant Communication from [34] (ITX = 4.8 mA, IRX = 3.4 mA, Iidle = 0.01 mA), Communication quality. With 16 Nodes, no MAC-level Fig. 6 shows that the model underestimates measured Iavg by packet loss occurred in either BCH mode (on or off); the 4.8–10.9×. Three factors drive this gap: (1) CPU overhead interference-free channel conditions produced no errors for from software-defined BCH decoding, CRC verification, and BCH to correct, so the coding gain could not be measured. frame parsing during RX; (2) Zephyr RTOS scheduling latency, Energy efficiency. Fig. 5 and Table VI show the measured which extends the active-radio window; and (3) MCU and current profile for one IBI. The SoC wakes briefly for TX peripheral activity during radio-off intervals (∼0.5 mA over and RX operations—visible as current spikes—and remains in 92% duty cycle). Substituting the measured state currents of low-power sleep for the rest of the interval (92% duty cycle, Table VI into (1) reduces the maximum absolute error to 7%, Table VI). Board-level currents exceed the nominal values; the confirming the model’s validity with implementation-measured RX overhead—dominated by software BCH decoding, CRCparameters. 16 verification, frame parsing, and RTOS scheduling latency— accounts for the largest gap. Fig. 6 compares measured Iavg with the analytical model of B. BLE Comparison Ruan et al. [38]; the gap is analyzed in Section VI-A. To quantify the determinism–efficiency tradeoff, we compared the SmartBAN implementation against BLE on identical hardware (nRF54L15, Zephyr BLE stack) with matched TABLE VI 1M PHY, CI = 220 ms, and slave latency 0 (Table VII). C URRENT C ONSUMPTION BY O PERATING S TATE Timing. SmartBAN’s TDMA assigns each Node a fixed slot State Mean ± Std Nominal Ratio Duty (%) within the IBI, so uplink arrivals depend only on the D-Beacon reference and remain time-aligned regardless of N (σ = 371 µs, TX (0 dBm) 5.83 ± 2.90 mA 4.8 mA 1.21× 1.02 Fig. 4). BLE, by contrast, schedules connection events through RXa 12.43 ± 7.96 mA 3.4 mA 3.66× 6.79 Radio offb 0.54 ± 1.28 mA — — 92.19 a centralized link layer whose contention grows with N [10]; arrival intervals fluctuate progressively as peripherals are added Peak (observed) 33.0 mA — — — (σ grows nearly tenfold from N =1 to N =16). Both protocols Node, 0 dBm TX, VDD = 1.8 V, approximately 10-minute meaachieve high PDR (SmartBAN >99.99%; BLE >99.9%). surement (628.9 s); values are mean ± std. dev. a RX comprises RX_Listen (14.10 ms, 94.4% of RX time) and Energy. SmartBAN consumes 4.4× more energy than BLE. RX_Decode (0.84 ms); the reported current is the dominant ListenHowever, the analytical model with nominal currents yields phase value. Duration-weighted average is 12.23 mA. (a) b Radio transceiver disabled; MCU active with Zephyr RTOS idle Iavg ≈ 289 µA—comparable to the BLE baseline (317 µA)— thread (sensor peripherals disabled in this measurement). indicating that the protocol-level duty cycle is not the dominant

PREPRINT

7

TX

16000

Active (28 ms)

12000

MCU Active

Sleep

InActive (184 ms)

Inactive Uplink TX

μA

D-Beacon RX

Slot2

RX_DCH avg (9846)

Radio OFF

8000 TX avg (6071)

8000

I-Ack RX wait

4000

4000

D-Beacon wait (8 ms)

Slot1

Slot0

12000

Current (μA)

RX (DCH)

0 0

5

10

15

20 Time (ms)

25

30

35

MCU/Sleep avg (1563/532 uA)

0 D0: IBI Active D1: TX D2: RX D3: Sleep 0

25

50

75

100

125

150

175

200

Time relative to IBI start (ms) Fig. 5. Measured current profile of one IBI (TD = 220 ms, Node). Current is color-coded by protocol state: TX (red), RX (blue), MCU active (gray), and sleep (green); the latter two together constitute the “Radio off” state in Table VI. Per-state averages are shown as dashed lines (corresponding to Table VI). The boxed region near the origin is magnified in the upper-right inset, where 10 ms slot boundaries are indicated by vertical dotted lines.

Average current Iavg (μA)

2500

Analytical (Ruan et al.) Corrected model Measured

2000

TABLE VII S AME -S O C P ROTOCOL C OMPARISON ( N RF54L15) Metric

BLE

99.997 1401 13.4

99.94 317 59.1

PDR (%)a Iavg (µA)b Battery life (d)c

1500 1000

SmartBAN

4.8× 6.2×

500

7.4×

10.9×

0 220

380

500

1000

193 B payload, 0 dBm TX, CI/IBI = 220 ms, 1 m indoor, 1 h each. a SmartBAN: 16 Nodes, 1-h aggregate (25-h PDR: 99.99%, Section V-D); BLE: N =1 Peripheral (1-h); BLE N =16 PDR is identical (99.94%). b PPK2 board-level, single Node/Peripheral; SmartBAN: 10-min window, BLE: 13-min window; BLE power is N -independent. c 450 mAh LiPo; battery-life extrapolation assumes steady-state operation.

IBI (ms)

asymmetry would favor BLE’s PDR. Fig. 6. Average current Iavg versus IBI. Markers: measured values with 95% confidence intervals. Solid line: analytical model of Ruan et al. [38] using nominal current values. Dashed line: corrected model using measured state currents from Table VI. The gap between the analytical curve and the measured data quantifies the system-level overhead inherent to a softwaredefined protocol stack on a general-purpose SoC.

factor. The difference between the measured 1401 µA and the model-predicted 289 µA, amounting to 1112 µA (79% of the total), is implementation overhead concentrated in the RX state (3.66× nominal, Table VI) from software-defined operations described in Section V-C. Channel diversity. SmartBAN operates on a single fixed data channel, whereas BLE employs adaptive frequency hopping across 37 data channels; in interference-rich environments, this

C. Limitations and Future Work Experimental scope. The experimental environment (Section V-A) was intentionally clean: no body shadowing, no co-channel interference, and a narrow RSSI spread of 20 dB. On-body deployment introduces time-varying multipath fading that may stress the timeout and reconnection mechanisms beyond what bench-top evaluation reveals. The implementation targets one SoC family (nRF54L15); although the software architecture defines a hardware-independent portability boundary (Section IV-A), porting to alternative platforms has not been verified. Model and protocol scope. The automaton decomposition maps naturally to TDMA-based protocols; applying the same

PREPRINT

8

approach to CSMA-based standards would require additional contention-resolution modeling. The first-fit slot allocation of Section IV-D is a standard heuristic, not a novel scheduling contribution. The presented implementation realizes operations (1), (2a), and (2b) of Table I; operations (3)–(6) are accommodated as bounded extensions of the behavioral model (Section III). Future work. On-body evaluation under dynamic channel conditions is the immediate priority. Other planned extensions include encryption, priority-based scheduling, and optional SmartBAN features (CAP, frequency hopping, relay [39]). VII. C ONCLUSION This paper presents the first operational SmartBAN realization—a complete PHY and MAC stack running on a commercial SoC (nRF54L15, Arm Cortex-M33, Zephyr RTOS). The standard left four behavioral omissions unspecified. Structured behavioral modeling addressed these by constructing Hub and Node automata (MH , MN ) as Mealy-type FSMs that serve as a design blueprint: phase control and connection lifecycle are captured at the model level, while slot-level timing and scheduling policy are resolved through structured realization on the target SoC. Experimental evaluation with sixteen concurrent sensor nodes over 25 h confirmed sub-millisecond timing jitter, >99.99% packet delivery, and autonomous disconnection recovery. A same-SoC BLE comparison characterized the determinism– efficiency tradeoff between the two protocols. The behavioral model established here provides the foundation for future extensions including on-body evaluation, encryption, downlink support, and priority-based scheduling. R EFERENCES [1] S. Movassaghi, M. Abolhasan, J. Lipman, D. Smith, and A. Jamalipour, “Wireless Body Area Networks: A Survey,” IEEE Commun. Surveys Tuts., vol. 16, no. 3, pp. 1658–1686, 2014. [2] L. Zhong et al., “Technological Requirements and Challenges in Wireless Body Area Networks for Health Monitoring: A Comprehensive Survey,” Sensors, vol. 22, no. 9, p. 3539, 2022. [3] ETSI, “Smart Body Area Network (SmartBAN); Low Complexity Medium Access Control (MAC) for SmartBAN,” European Telecommunications Standards Institute, ETSI TS 103 325, Jul. 2022. [4] ——, “Smart Body Area Network (SmartBAN); Enhanced Ultra-Low Power Physical Layer,” European Telecommunications Standards Institute, ETSI TS 103 326, Jul. 2021. [5] IEC, “Smart body area network (SmartBAN) – Enhanced ultra-low power physical layer,” International Electrotechnical Commission, IEC 63203-801-1, 2022. [6] ——, “Smart body area network (SmartBAN) – Low complexity medium access control (MAC) for SmartBAN,” International Electrotechnical Commission, IEC 63203-801-2, 2022. [7] T. Paso et al., “An Overview of ETSI TC SmartBAN MAC Protocol,” in Proc. 9th Int. Symp. Med. Inf. Commun. Technol., 2015, pp. 10–14. [8] Bluetooth SIG, “Bluetooth Core Specification, Version 6.2,” https://www. bluetooth.com/specifications/specs/core-specification/, 2025. [9] C. Gomez, J. Oller, and J. Paradells, “Overview and Evaluation of Bluetooth Low Energy: An Emerging Low-Power Wireless Technology,” Sensors, vol. 12, no. 9, pp. 11 734–11 753, 2012. [10] P. H. Kindt, D. Yunge, R. Diemer, and S. Chakraborty, “Energy Modeling for the Bluetooth Low Energy Protocol,” ACM Trans. Embed. Comput. Syst., vol. 19, no. 2, pp. 13:1–13:32, Mar. 2020. [11] IEEE, IEEE Standard for Local and Metropolitan Area Networks—Part 15.6: Wireless Body Area Networks, Std. IEEE 802.15.6-2012, Feb. 2012. [12] M. Yoshimi, T. Seike, T. Ito, H. Shibamura, and H. Tanaka, “Relationship between RSSI and Human Posture in Wireless Body Area Network,” in Proc. IEEE 13th Global Conf. Consumer Electron., 2024, pp. 826–827.

[13] M. Yoshimi, T. Ito, and H. Tanaka, “Multi-Node SmartBAN Implementation for Wearable IMU Sensors,” in Proc. 48th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., 2026, accepted, poster. [14] A. M. El-Mohandes, A. Shalaby, and M. S. Sayed, “Robust Low Power NB PHY Baseband Transceiver for IEEE 802.15.6 WBAN,” in Proc. IEEE 27th Int. Conf. Microelectron. (ICM), 2015, pp. 91–94. [15] X. Su, C. Li, and X. Yuan, “IEEE 802.15.6-Based Prototype System for Wireless Body Area Network: Design and Implementation,” arXiv:1701.02421 [cs.NI], 2017. [16] C. Bachmann et al., “A 3.5 mW 315/400 MHz IEEE 802.15.6/Proprietary Mode Digitally-Tunable Radio SoC with Integrated Digital Baseband and MAC Processor in 40 nm CMOS,” in Proc. Symp. VLSI Circuits (VLSIC), 2015, pp. C94–C95. [17] L. Leonardi, G. Patti, and L. Lo Bello, “Multi-Hop Real-Time Communications Over Bluetooth Low Energy Industrial Wireless Mesh Networks,” IEEE Access, vol. 6, pp. 26 505–26 519, 2018. [18] R. Ben Hamouda and I. Ben Hafaiedh, “Formal modeling and verification of a wireless body area network (WBAN) protocol: S-TDMA protocol,” in Proc. Int. Conf. Internet Things, Embedded Syst. Commun. (IINTEC), Oct. 2017, pp. 72–77. [19] B. Touijer, Y. Ben Maissa, and S. Mouline, “IEEE 802.15.6 CSMA/CA access method for WBANs: Performance evaluation and new backoff counter selection procedure,” Comput. Netw., vol. 188, p. 107759, Apr. 2021. [20] M. Hämäläinen et al., “Recent Progress in ETSI TC SmartBAN Standardization,” in Proc. 17th IEEE Int. Symp. Med. Inf. Commun. Technol. (ISMICT), May 2023, pp. 1–6. [21] H. Viittala, L. Mucchi, M. Hämäläinen, and T. Paso, “ETSI SmartBAN System Performance and Coexistence Verification for Healthcare,” IEEE Access, vol. 5, pp. 8175–8182, 2017. [22] H. Viittala, M. Hamalainen, and L. Mucchi, “Performance of the ETSI SmartBAN System in the Interfered IEEE 802.15.6 Channel,” in Proc. IEEE Int. Symp. Med. Inf. Commun. Technol., 2017, pp. 15–18. [23] R. Khan, M. M. Alam, T. Paso, and J. Haapola, “Throughput and Channel Aware MAC Scheduling for SmartBAN Standard,” IEEE Access, vol. 7, pp. 63 133–63 145, 2019. [24] R. M. Khan and M. M. Alam, “SmartBAN Performance Evaluation for Diverse Applications,” in Body Area Networks: Smart IoT and Big Data for Intelligent Health Management (BODYNETS 2019), ser. LNICST, vol. 297. Springer, 2020, pp. 239–251. [25] R. Khan, M. M. Alam, and M. Guizani, “A Flexible Enhanced Throughput and Reduced Overhead (FETRO) MAC Protocol for ETSI SmartBAN,” IEEE Trans. Mobile Comput., vol. 21, no. 8, pp. 2671–2686, Aug. 2022. [26] L. Ruan, M. P. I. Dias, and E. Wong, “Time-Optimized MAC for Low Latency and Energy-Efficient SmartBANs,” in Proc. IEEE INFOCOM Workshops, 2016, pp. 1089–1090. [27] J. Ramis-Bibiloni and L. Carrasco-Martorell, “Lengthening Battery Life Expectancy of Sensors in WBANs: A Multifactorial Approach,” Internet of Things, vol. 25, p. 101071, 2024. [28] A. D’Souza, H. Viittala, M. Hämäläinen, and L. Mucchi, “Performance Comparison Between ETSI SmartBAN and Bluetooth,” in Proc. IEEE Int. Symp. Med. Inf. Commun. Technol., 2018, pp. 1–5. [29] G. H. Mealy, “A Method for Synthesizing Sequential Circuits,” Bell Syst. Tech. J., vol. 34, no. 5, pp. 1045–1079, 1955. [30] J. E. Hopcroft, R. Motwani, and J. D. Ullman, Introduction to Automata Theory, Languages, and Computation, 3rd ed. Pearson, 2006. [31] D. Harel, “Statecharts: A Visual Formalism for Complex Systems,” Sci. Comput. Program., vol. 8, no. 3, pp. 231–274, Jun. 1987. [32] D. Harel and A. Naamad, “The STATEMATE semantics of statecharts,” ACM Trans. Softw. Eng. Methodol., vol. 5, no. 4, pp. 293–333, Oct. 1996. [33] Seeed Studio, “XIAO nRF54L15,” https://wiki.seeedstudio.com/xiao_ nrf54l15/, 2025. [34] Nordic Semiconductor, nRF54L15, nRF54L10, and nRF54L05 Wireless SoCs — Datasheet v1.0, https://docs.nordicsemi.com/bundle/ps_ nrf54L15/, Nordic Semiconductor ASA, sep 2025. [35] Zephyr Project, “Zephyr RTOS,” https://docs.zephyrproject.org/, 2025. [36] Nordic Semiconductor, “nRF Connect SDK,” https://www.nordicsemi. com/Products/Development-software/nRF-Connect-SDK, 2025. [37] ——, “Power Profiler Kit II (PPK2) User Guide v1.0.1,” https://docs. nordicsemi.com/bundle/ug_ppk2/, 2024. [38] L. Ruan, M. P. I. Dias, and E. Wong, “SmartBAN with Periodic Monitoring Traffic: A Performance Study on Low-Delay and High Energy-Efficiency,” IEEE J. Biomed. Health Inform., vol. 22, no. 2, pp. 471–482, Mar. 2018. [39] ETSI, “Smart Body Area Network (SmartBAN); MAC Extension with Relay Capability,” European Telecommunications Standards Institute, ETSI TS 103 805, Jan. 2024.

Related documents

• Cybersecurity Law and Regulation in China: Securing the Smart State#671916
• Binding the Smart City Human-Digital System with Communicative Processes#100205
• SELF-HEALING RADIATION-DAMAGED ELECTRONICS VIA INTEGRATED ON-CHIP MICROHEATERS#657692
• Protocol effects on feature-based hardware-Trojan detection across Trust-Hub families#673624
• Commodification, datafication and smart cities: An ethical exploration#99815