Bluetooth Mesh Stack

The Bluetooth Core Specification [ 1 ] includes a low energy version for low-rate wireless personal area networks, referred as Bluetooth low energy (BLE) or Bluetooth Smart. The BLE stack consists of the: generic attribute profile (GATT), attribute protocol (ATT), security manager protocol (SMP), logical link control and adaptation protocol (L2CAP), link layer (LL) and physical layer (PHY). BLE was added to the standard for low energy devices generating small amounts of data, such as the notification alerts used in applications like home automation, healthcare, fitness, and the Internet of Things (IoT).

The Bluetooth Mesh Profile [ 2 ] defines the fundamental requirements to implement mesh networking solutions for BLE. The mesh stack is located on top of the Bluetooth Core Specification and consists of the: model layer, foundation model layer, access layer, upper transport layer, lower transport layer, network layer and bearer layer. Bluetooth mesh networking enables end-to-end communication in large-scale networks to support applications like smart lighting, industrial automation, sensor networking, asset tracking, and many other IoT solutions.

Mesh Stack

This figure shows the Bluetooth mesh stack over the advertising bearer.

BLE Core Stack

This example models these layers of the BLE core stack:

Discrete Event Simulation

DES is a type of simulation that models the functioning of a system as a discrete sequence of events in the time domain. Each event occurs at a specific time epoch and subsequently marks a change of state in the system. As a result, the simulation can directly jump from event to event in the time domain. The fundamental advantages of using DES in this example are:

Check for Support Package Installation

% Check if the 'Communications Toolbox Library for the Bluetooth Protocol'
% support package is installed or not.
commSupportPackageCheck('BLUETOOTH');

Multi-Node Bluetooth Mesh Network Model

This example models a Bluetooth mesh network with 21 nodes. The model outputs PDR of the network along with different statistics such as the number of transmitted, received, and dropped packets at physical, link, and network layers, and also a plot visualizing the network scenario. The modeling includes:

To configure a specific scenario, do one of these:

Bluetooth node

Each node is modeled as a subsystem with a network stack, which includes the Bluetooth mesh packet generator and receiver, network layer, LL, and PHY.

Application layer

The application layer is implemented to generate and receive application traffic. It is divided into two sub-blocks:

Network layer

The network layer is modeled as a DES block. This block is responsible for transmitting the lower transport layer messages over the advertising bearer and relaying the mesh messages when the 'Relay' feature is enabled. When a network PDU is received, this block decodes the received PDU. If the PDU is decoded successfully, then the decoded information is passed to the lower transport layer.

You can configure the relay feature, network transmit interval, network transmit count, relay retransmit interval, and relay retransmit count by using mask parameters of the Network layer block.

Link layer

The link layer is modeled as a DES block. This block maintains a state machine for LL Broadcaster and Observer roles. This block is responsible for transmitting and receiving the mesh advertising packets by using bleLLAdvertisingChannelPDU and bleLLAdvertisingChannelPDUDecode functions.

You can configure scan and advertising intervals by using mask parameters of the Link layer block.

Physical layer

The PHY functionality includes:

LL initiates packet transmission by sending an LL packet and Tx indication to the PHY Tx block. This block generates a waveform for the received LL packet by using the bleWaveformGenerator function. It also scales the samples of the BLE waveform with the configured Tx power (assuming Tx gain is 0). The generated BLE waveform is transmitted through the shared channel. The shared channel is modeled by using the SimEvents Multicast Queue.

You can configure the Tx power (dBm) by using mask parameters of the PHY Tx block.

The free-space path loss model is added to the transmitted BLE waveform as channel impairments. You can choose to enable or disable this impairment. In addition to this impairment model, the signal reception range can also be limited by using an optional range propagation loss model. To model any of these channel impairment options, the channel model must contain the position of both the sender and the receiver. The channel is modeled inside each receiving node, before passing the BLE waveform to the PHY Rx block.

You can configure channel impairments by using mask parameters of the BLE channel block.

This block applies thermal noise and interference to the received BLE waveform (assuming Rx gain is 0). Thermal noise is modeled by using the comm.ThermalNoise function with the configured value of the noise figure. Interference is modeled by adding the IQ samples of both the interfered and the actual signals. After applying thermal noise and interference, PHY Rx block decodes the resultant waveform. If the LL packet is decoded successfully, then it is passed to the LL.

You can configure the noise figure (in dB) using mask parameters of the PHY Rx block.

Node position allocator (NPA) Assigns the location of nodes in the mesh network. This block supports linear, grid, and list position allocation strategies.

Visualizer This block is used to visualize the mesh network scenario in the simulation. You can configure this block to visualize the specified configuration. You can enable or disable visualization by using the mask parameters of this block.

Simulation Results

The results obtained in this simulation are:

The PDR is the ratio of number of received packets at the destination to the number of packets transmitted by the source and is given by:

This model outputs PDR for this multi-node mesh network and is saved to a base workspace variable named PDR.

This model outputs statistics of each node in the workspace variable statisticsAtEachNode. The statistics captured at each node are:

A plot with visual representation of the mesh network scenario is shown in the simulation. You can see the statistics of each node by placing your cursor over it.

This example shows how to configure and simulate a multi-node Bluetooth mesh network by using DES. The mesh network model in this example outputs PDR as a workspace variable with a visual representation of the mesh network.

Further Exploration

To observe the variation in the network PDR, you can vary the configuration parameters at the mesh packet generator, the network layer, LL and PHY. In these simulation results, you can see the impact of network layer repetitions (NLR) on the network PDR.

The NLR includes the repetitions of both the network messages and the relayed messages. The working principle of flood-based networks ensures that the message reaches the destination node. Therefore, it is important to retransmit the network and the relay messages. The number of NLR is dependent on the network configuration of the given network topology. Increasing the number of NLR ensures that the likelihood of the messages reaching the desired destination node is high. However, specifying a high value of the NLR can have adverse effects on the network performance parameters such as the overhead, energy consumption, and the duty cycle. As a result, it is essential to tune the value of NLR for a given network topology and achieve an efficient tradeoff between the PDR and network performance.

In the preceding figure you can see that the PDR increases with the NLR and decreases with the number of source nodes in the network. For a specific value of the NLR, the PDR value reaches 1 and thereafter it stabilizes. This specific value of the NLR might vary based on the network configuration parameters such as the total number of nodes, location of the nodes, number of source nodes, number of relay nodes, and so on. You can run helperBLEMeshDESPDRCalculation to reproduce these results by using three source nodes. Set the number of source nodes to two and five to get the corresponding results. You can run the simulations for any custom network scenario and get the optimal value of the NLR.

Apart from the NLR, the PDR varies with respect to multiple configuration parameters stated in helperBLEMeshDESPDRCalculation. You can further explore the mesh network model by varying any of these parameters.

Appendix

The example uses these features:

The example uses these helpers:

References