BUBBLE-BLUE: Autonomous Bluetooth Mesh Network

• BUBBLE-BLUE is an initiative to create autonomous, secure ad hoc Bluetooth mesh networks using smartphone-based pico-nets without reliance on centralized infrastructures.
• It employs dynamic Connected Dominating Sets for efficient routing and robust AES-based symmetric-key cryptography to ensure message confidentiality and member revocation.
• The project supports both civilian and military applications, providing resilient communication in emergencies and high-risk scenarios as validated by empirical simulations.

The BUBBLE-BLUE (BB) Project is an initiative to realize autonomous, secure, and resilient multihop private networking using Bluetooth on commodity smartphones. It advances a “terrestrial STARLINK” paradigm, leveraging the global proliferation of smartphones to create a distributed ad hoc Bluetooth mesh network capable of functioning independently of cellular or data infrastructures. BB’s architecture exploits dynamic Connected Dominating Sets (CDS) for scalable routing, incorporates robust symmetric-key cryptography for privacy, and is designed for both civilian and military applications—enabling messaging, multimedia sharing, live topology discovery, and secure field operations in environments where traditional communication channels are unavailable or compromised.

1. Project Objectives and System Overview

BUBBLE-BLUE defines a terrestrial mesh network formed dynamically by smartphones using only their built-in Bluetooth interfaces. Each BB “bubble” is a localized, private pico-network formed and maintained autonomously, without cellular or wide-area Internet support. The overall vision is to create a resilient, stealthy alternative to satellite-based communication networks (e.g., STARLINK), with applications in emergency/disaster response, military operations, and dense environments where operator networks are inaccessible or overloaded.

The dual-use architecture serves:

• Civilian users: providing chat, geolocation, and media transfer with local autonomy.
• Military/secure users: offering encapsulated AES key management and member reputability for enhanced confidentiality and integrity.

No reliance on centralized infrastructure or mobile operator support is assumed at any stage.

2. Network Formation and Security Mechanisms

Private Bluetooth Bubbles and Pico-Networks

Each smartphone acts as a Bluetooth node, participating in multiple overlapping pico-nets. Pico-nets are self-organized with randomly chosen master nodes and multiple slave nodes. Overlapping memberships are permitted, enabling the emergence of continuous coverage and robust local topologies.

Key features:

• A device can act as a bridge between bubbles by participating in multiple pico-nets.
• Neighborhood is defined through shared pico-nets; nodes sharing at least one pico-net exchange control and user data.

Cryptographic Key Management and Communication

BB implements a pairwise pre-distributed symmetric key model using AES for confidentiality:

• At network initialization, an $n \times n$ symmetric key matrix $K_{i, j}$ ($i, j \in \{1, ..., n\}$) is generated by a group leader.
• Each message is encrypted with a one-time session key $K$.
• For every recipient $i \neq \text{originator}$, the header encodes $K$ as $K_i = AES(K_{i, j}, K)$.
• The sender’s field is set to a fixed dummy value (all bits to 1) to signal originator status.

This construct ensures:

• Only legitimate bubble members can recover the session key and decrypt content.
• Compromised members can be effectively revoked by setting their key field to zero in outgoing messages.

Each message includes both header (encoded session keys) and encrypted payload (subfields for originator, type, and data).

Control and Data Messaging

The BB protocol stack specifies two major message families:

• Control messages: periodic “hello” for neighbor discovery, topology updates, ARQ retransmission logic, mute/failure notifications, and group (bubble) management.
• Information messages: user data including chat, media, or geolocation sharing.

All control interactions are autonomously self-managed based on local topology evolution and group changes.

3. Smartphone-Based Terrestrial STARLINK Topology

The BB architecture leverages billions of smartphones as both endpoints and routing relays:

• Bluetooth is used exclusively—offering both stealth (RF profile avoids common spectrum surveillance) and energy efficiency, suitable for prolonged off-grid use.
• The system is specifically targeted at compromised, congested, or high-risk areas wherein traditional networks fail or must be circumvented.
• Network robustness is directly proportional to smartphone density; high global proliferation ensures broad geographic coverage, unlike high-cost satellite constellations.

Practical use cases span emergency rescue, field coordination during crises, secure military deployment, and large public events.

4. Routing Optimization: Dynamic Connected Dominating Sets

Motivation and Model

Bluetooth’s lower layers only support unicast, so all-broadcast must be virtualized through judicious selection of forwarding sets. BUBBLE-BLUE adopts dynamic Connected Dominating Set (CDS) approaches for minimal, robust broadcast coverage.

CDS Algorithm Variants

• Wu–Li (1999): Each node uses two-hop neighborhood knowledge (gathered via hello packets); inclusion rules depend on identifier ordering and local connectivity (a node joins CDS if its removal partitions the local subgraph).
• MPR (Multi-Point Relay) based CDS: Nodes select a minimal subset (MPR set) for effective two-hop broadcast; this approach yields smaller CDSs but increases per-node selection computations.
• Additional candidates (e.g., Wu–Li 2001, covering tree solutions) are discussed, each with explicit tradeoffs regarding CDS size, election overhead, and broadcast delay due to Bluetooth’s serialized unicast transmission semantics.

CDS Broadcast Mechanism

• Node broadcasts are propagated by CDS members only; all duplicates are suppressed by sequence number tracking to avoid redundant forwarding.
• Topology control is maintained by periodic neighbor set broadcasts from CDS members, allowing global topology reconstruction.

Cost Analysis and Mathematical Formalism

• For a graph $G = (V, E)$ and a CDS $V’ \subset V$:

$$\text{Avg. Broadcast Cost} = \frac{2|E|}{n} + \left(1 - \frac{1}{n} \right) \operatorname{deg}(V’)$$

where $|E|$ is the edge count and $\operatorname{deg}(V’)$ is the sum of the degrees of CDS nodes.

• Simulations (1D and 2D unit disk graphs) give density-dependent cost formulas; in 1D high-density models, average densities are 1.0 (MPR flooding CDS), 1.5 (MPR CDS), and 2.0 (Wu-Li CDS), translating into total transmission requirements as functions of node density $\lambda$ and network length $\ell$, e.g.

$\text{MPR flooding:} \quad 2(\ell + 1)\lambda - 2$

Wu-Li strategies maintain a favorable cost/coverage tradeoff without requiring full network synchronization.

Integer Programming Model for Optimal CDS

Define binary variables $x_i$ for node inclusion, and auxiliary flow variables $f_{i,j}$ for connectivity:

• Minimize $\sum_i \deg(i) x_i$, subject to

$$x_i + \sum_{j \in N(i)} x_j \geq 1 \quad \forall\, i$$

and connectivity is enforced by a single-commodity flow rooted at an arbitrary node $r$:

$$\sum_{j:(j,i) \in E} f_{j,i} - \sum_{j:(i,j) \in E} f_{i,j} = \begin{cases} k-1 & \text{if } i = r \ -1 & \text{if } x_i = 1,\, i \neq r \ 0 & \text{if } x_i = 0 \end{cases}$$

with link capacities $f_{i,j} \leq (|V|-1)x_i$.

5. Simulation Results and Empirical Validation

The methodology has been validated by geometric graph simulations in both one and two dimensions. Key empirical results include:

• In 1D high-density regimes, the Wu–Li CDS achieves sufficient coverage with a cost penalty of 2 messages per node, outperforming MPR CDS in tested scenarios.
• 2D simulations confirm similar behavior, verifying that the CDS approach is scalable and robust, with only moderate overhead compared to theoretical minima.
• These analyses facilitate operational parameter selection for implementation under both civilian and tactical conditions, balancing power, resilience, and communication latency.

6. Impact, Limitations, and Practical Significance

The BUBBLE-BLUE architecture addresses critical communication gaps in denied, degraded, or disconnected environments:

• Power-efficient, infrastructure-free deployment is realized via the existing global smartphone base.
• AES-based key management enables strong per-message privacy and revocation.
• Dynamic CDS construction ensures scalable, loop-free broadcast even in ad hoc, mobility-prone scenarios.

Potential limitations include:

• Scalability constraints with large $n$ due to the $n \times n$ key matrix and header size, though this is tractable for tactical or localized use cases.
• Transmission delays inherent in multi-unicast Bluetooth forwarding; however, the CDS strategy significantly mitigates redundant traffic.

Future enhancements could explore adaptive CDS election under mobility, more granular revocation models, and optimization for asymmetric Bluetooth link properties.

---

The BUBBLE-BLUE project provides a technically grounded framework for building terrestrial, infrastructure-independent multihop Bluetooth networks. It merges established graph-theoretical broadcast optimizations with robust symmetric cryptography, exploiting ubiquitous hardware to deliver secure, ad hoc communications with quantifiable cost and coverage assurances (Achir et al., 31 Aug 2025).