Access Official Technical PDF

Share Summary:LinkedInEmail

Bluetooth is widely available in connected devices, but its traditional strength has been short-range, single-hop communication. Concurrent transmissions Bluetooth techniques offer another possibility: multiple nodes send the same packet at nearly the same time, allowing nearby receivers to recover the message and forward it across a network. BlueFlood applies this approach to Bluetooth 5 advertising, demonstrating a route to network-wide dissemination without relying on conventional per-link forwarding.

For industrial automation teams, the idea is relevant wherever Bluetooth-enabled devices need to distribute the same information across multiple hops. It also comes with an important qualification: successful concurrent transmission depends on close timing and favorable radio conditions. The approach is a useful networking primitive, not a guarantee that overlapping packets will always decode.

How concurrent transmissions work

Conventional wireless protocols generally try to prevent simultaneous transmissions because overlapping signals can cause collisions. Concurrent transmission (CT) takes a different approach. Several nodes transmit identical data in close synchronization. A receiver may then decode the packet despite the overlapping signals, through capture effects and non-destructive interference.

The signals do not necessarily combine into a stronger, clean version of the original. Differences in transmitter timing, received signal strength, carrier frequency and phase distort the combined waveform. Yet if the packets carry the same content and arrive sufficiently close together, the distortion may still allow the receiver to recover the message. This is distinct from sending different data at once, where a sufficiently strong signal may need to dominate the others for capture to occur.

That distinction matters for network design. A flood of one shared message can make use of identical-packet transmissions. Simultaneously combining different sensor readings or payloads is a harder problem and is not what BlueFlood demonstrates.

Why Bluetooth presents a different challenge

Bluetooth Low Energy (BLE) uses Gaussian Frequency Shift Keying (GFSK), while 2.4 GHz IEEE 802.15.4 uses O-QPSK and direct-sequence spread spectrum (DSSS). DSSS adds redundancy that can help a receiver handle signal distortion. Bluetooth does not use DSSS, so the behavior of established concurrent-transmission approaches for 802.15.4 cannot simply be assumed to carry over.

Bluetooth 5 provides four physical-layer modes: uncoded 1 Mbps and 2 Mbps, plus coded long-range modes at 500 Kbps and 125 Kbps. The coded modes use forward error correction (FEC), adding redundancy that may improve resilience, but at the cost of a lower data rate. The uncoded modes offer higher rates without that protection.

Concurrent signals also differ slightly in carrier frequency because transmitters have independent oscillators. When signals at slightly different frequencies overlap, their combined amplitude rises and falls in a pattern known as beating. These variations can create intervals of weaker reception and may cause errors to occur in bursts. Timing offsets can further distort the symbols; the tolerable offset is especially tight at the 2 Mbps rate because its symbol period is shorter.

Modeling and testing the Bluetooth physical layer

BlueFlood’s evaluation first examines whether concurrent transmission is physically feasible with frequency-shift keying. The analysis models how overlapping signals affect the received waveform and the bit error rate (BER). It considers factors including relative timing, signal power and carrier-frequency offset. Because Bluetooth uses GFSK, the model uses binary FSK as a simpler related modulation scheme rather than treating the two as identical.

The model highlights a trade-off: two synchronized transmissions can provide an advantage in some noisy conditions, but the combined signal can perform worse than a single transmission in other conditions. In particular, beating can create periods of destructive interference, and more concurrent transmitters can produce a more complex signal envelope. The authors therefore describe CT as feasible but fragile, with results dependent on the specific radio and transmission conditions.

Controlled experiments on Bluetooth radios are used to validate the analysis and compare Bluetooth 5 modes under different signal powers, timing offsets and carrier-frequency offsets. This step is significant because practical performance cannot be inferred from an ideal model alone. Radios have real oscillator differences, and software and hardware must coordinate packet timing closely.

BlueFlood turns feasibility into multi-hop dissemination

BlueFlood uses the CT principle to spread information across a multi-hop Bluetooth network. Its design follows the basic flooding pattern: a designated initiator sends a packet, nodes that receive it join the dissemination by retransmitting the same information, and the process carries the message outward. Tight synchronization helps those retransmissions overlap in a way that receivers can tolerate.

The approach is particularly compatible with Bluetooth advertising. Advertising packets are broadcasts rather than messages in a connected, device-to-device session, and they can be received by standard Bluetooth devices. In the evaluation, BlueFlood’s beacons could be received by off-the-shelf smartphones, without requiring specialized receiver hardware or software. That offers a way to make a network-distributed message visible to existing devices, although it does not mean smartphones participate in forwarding the flood.

In tests conducted in two university-building testbeds, BlueFlood achieved a 99.9% end-to-end delivery ratio for periodic dissemination of 38-byte advertising packets. Packets were sent at 200-millisecond intervals using the 2 Mbps mode, with a reported duty cycle of 0.4%. These results demonstrate the protocol under the evaluated conditions; they should not be read as a universal performance guarantee for every building, radio configuration or interference environment.

Practical considerations for automation teams

For system integrators and plant engineers, CT-based dissemination may be worth evaluating when many devices need to receive the same short message. Potential examples include broadcasting a status or alert across a deployment, or supporting flooding-based communication in a Bluetooth Mesh context. The central design advantage is sender diversity: multiple nodes can help deliver the same information without requiring a separately maintained route for each receiver.

Deployment decisions should account for the method’s constraints:

  • Timing precision: retransmissions need to be closely aligned; tighter timing becomes more challenging at higher symbol rates.
  • Radio variation: carrier-frequency offsets and signal-power differences influence the combined waveform and reception quality.
  • Payload and traffic pattern: the reported delivery result applies to the evaluated 38-byte packets and 200-millisecond interval.
  • Network conditions: testbed results provide evidence of feasibility, but local validation is needed before relying on a particular delivery target.

Conclusion

BlueFlood shows that Bluetooth 5 can support multi-hop dissemination based on concurrent transmissions, despite the challenges posed by GFSK, limited physical-layer redundancy in uncoded modes, and imperfect synchronization. Its building testbeds demonstrate high delivery for a specific periodic advertising workload, while the analysis makes clear that overlapping transmissions can also reduce reception reliability. For automation decision-makers, the practical takeaway is to consider CT as a promising flooding technique for shared messages—and to validate timing, radio conditions and required reliability in the intended deployment.

Access Official Technical PDF

About The Author