Understanding LoRa Physical Layer: PHY Modulation in LoRaWAN

The physical layer is the foundation of wireless communication, defining how data is transmitted over the air. It includes parameters such as frequency selection, modulation schemes, and power settings. Understanding these attributes is crucial for configuring LoRaWAN devices to operate effectively in different environments and under various regulatory constraints.

In this article, we will explore the LoRa physical layer, covering modulation schemes, spreading factors, frequency bands, and other key attributes used in the LoRaWAN network stack at layer-1 of the OSI model.

Introduction

As we know, LoRaWAN is one of the LPWAN technologies. It offers wide coverage and long range in addition to low power consumption.

There are different frequency bands allocated throughout the world for this wireless system. It includes the 902 to 928 MHz frequency band in the US, 863-870 MHz/433-434 MHz in Europe, and 470-510 MHz/779-787 MHz in China.

The figure depicts the protocol stack layers of a LoRaWAN system. As shown, the LoRa physical layer lies between the RF and MAC layers.

Following are the functions of the LoRa PHY Layer:

• It offers services to upper layers above it i.e., the MAC layer and Application layer.
• It passes the physical layer frame to the radio frequency layer for up-conversion to radio frequency as per region such as US, EU, China, Australia, INDIA, and so on.
• It uses Forward Error Correction to help in error correction.
• It adds a preamble to the physical layer frame which is used in time/frequency synchronization at the receiver.
• It uses CRC to help in error detection of the PHY frame carrying the MAC payload.
• It uses modulation such as GFSK to make an analog signal suitable to be transmitted over the air.

LoRa Physical Layer Modulation

Before we delve into the physical layer, let’s understand some LoRaWAN terminologies.

Let’s understand the LoRa Physical Layer modules:

• The Physical Layer constructs the frame in order to transmit the payload from the MAC layer over the RF link.
• It inserts PHDR, PHDR_CRC, preamble, and CRC for the entire frame. The CRC field is available in the uplink message only.
• LoRa uses wideband linear frequency modulated chirp pulses to encode information bits. A chirp is called a sweep signal which is a tone in which frequency increases with time (called an up-chirp) or decreases with time (called a down-chirp). In an up-chirp the frequency increases in time, and in a down-chirp the frequency decreases in time. These chirp signals are used as carrier signals where a message is encoded on.

Following figure depicts encoding of message on chirp signals. As shown starting part is preamble sequence followed by two down chirps and encoded payload.

The figure depicts a LoRa modulated message.

The chirps are cyclically shifted, and it is the frequency jumps which determine data encoding on the chirps.

Let’s understand the use of spreading factor in LoRa to encode a message.

Example: A symbol consisting of 1011111 (representing a decimal of 95). Here the number of raw bits to be encoded by one symbol is 7. Hence, Spreading Factor = 7.

The symbol has $2^{SF}$ values. For SF=7, we have a range from 0 to 127. The symbol value is encoded on a sweep signal (up-chirp) i.e., the sweep signal is divided into $2^7 = 128$ chips.

This is shown in the figure below.

• The PHY layer passes the modulated signal to the RF transceiver for up-conversion to higher frequencies as per country-specific LoRaWAN bands. Similarly, down-conversion is carried out at the other end before demodulation to raw information bits.

As mentioned above, the spreading factor defines two values:

• The number of raw bits which can be encoded by that symbol and
• Each symbol can hold $2^{SF}$ chips.

Forward Error correction is applied to payload of the loRa packet. In this process, error correction bits are added to the transmitted data. This added redundant bits help to restore the data when they get corrupted by interference. More the redundant bits, more easier it helps to correct the corrupted data but at the cost of bandwidth and reduction in battery life. LoRa uses different coding rate values such as 4/5, 4/6, 4/7 and 4/8.

$CR = 4/(4+CR)$, Where $CR = 1,2,3,4$

Following LoRaWAN physical layer parameters are configured:

• Bandwidth (BW) = 125 KHz to 500 KHz
• Spreading Factor (SF) = 7 to 12
• Coding Rate (CR) = 1 to 4

In LoRa, Bandwidth is used interchangeably with chip rate, hence BW = Rc = Chip rate (Chips/sec)

$Rc$ (symbols/sec) = $BW/2^{SF} = Rc/2^{SF}$

For example, if BW = 125 KHz and SF = 7

$R_s = 125000/2^7 = 977$ Symbols/sec.

$R_b$ (bits/sec) = $SF * (BW/2^{SF}) * (4/(4+CR))$

The figure depicts the LoRaWAN physical layer MATLAB simulation model.

Reference: LoRaWAN specifications published by LoRa Alliance ( https://www.lora-alliance.org ).

Conclusion

Mastering the characteristics of the LoRa physical layer allows users to optimize their network for coverage, power consumption, and interference management. The flexibility of the physical layer parameters makes LoRaWAN an adaptable solution for a wide range of IoT applications. By fine tuning these settings, users can enhance the performance and efficiency of their LoRaWAN deployments.

MATLAB code for FSK and GFSK used in LoRaWAN/LoRa