Secure Internet of Things Device with Single-Channel Communication for Resource-Constrained Applications ^†

Abstract

Internet of Things (IoT) devices are being utilized in large numbers in various applications, ranging from healthcare, manufacturing, and home automation to manufacturing, etc. This rapid increase in the number of devices has also led to a significant increase in cybersecurity concerns for these devices. Malicious nodes exploit vulnerabilities in conventional IoT architectures, leading to high risks to data integrity, privacy, and system reliability. Hence, there is a need for innovative solutions to improve the security of IoT devices and networks. Considering this, in this paper, a novel approach towards securing IoT devices against any malicious attack is proposed. A new IoT node is proposed, that has a single communication channel, which sends information only to a specific endpoint. Unlike a conventional IoT system with multiple communication channels, the proposed design limits communication to a single dedicated device, thereby drastically reducing the attack probability. The proposed device architecture integrates a robust cryptographic protocol to establish a secure and authenticated communication link between the IoT device and its designated endpoint. By employing a state-of-the-art encryption technique and secure access controls, the proposed solution can mitigate common attacks such as eavesdropping, data tampering, and unauthorized access. The proposed system also improves resource efficiency, lowers the device’s power consumption due to single-communication-channel data transmission, and simplifies network management, thus making it suitable for resource-constrained applications, such as military surveillance, home automation, etc.

1. Introduction

The Internet of Things (IoT) has become an integral part of everyday life, with applications ranging from smart homes to healthcare systems, industrial automation, automobiles, transportation and logistics, defense systems, etc. This growth of IoT over the years (Figure 1) depicts the ubiquity of IoT devices, which has led to a high level of connectivity, enabling real-time data collection, monitoring, and control [1,2]. But with this increased connectivity, significant security issues also arise. IoT devices are deployed in environments wherein they interact with sensitive data, such as the personal information of users, financial details and transactions, infrastructure controls, etc. The potential consequences of security breaches in these contexts can be critical, and can include privacy violations, financial loss, and even threats to human safety [3,4,5,6].

Due to the limited computational resources and capability of IoT devices, the traditional security measures such as firewalls, antivirus, intrusion detection systems, etc., are not always suitable. Further, the distributed nature of IoT networks creates a diverse attack pool and diverse attack possibilities, thereby making it lack a uniform security solution for all the attacks. Additionally, cyberattacks on IoT devices exploiting weak authentication and unencrypted communication channels have also increased in numbers.

Hence, there is a need for developing security solutions specifically designed for IoT environments. These solutions must be lightweight and efficient to protect against different threats such as tampering, data breaches, and unauthorized access. Considering these constraints, this study proposes a system and a method for securing sensing devices in resource-constrained IoT applications, with a particular focus on preventing unauthorized access and ensuring the integrity of the sensed data.

The rest of the paper is organized as follows: Section 2 presents a review of various security issues and factors limiting power management in IoT devices. The proposed system architecture and methodology are presented in Section 3, followed by a discussion in Section 4 and the conclusion in Section 5.

2. Literature Review

The IoT architecture basically consists of four layers, which are the perception layer, the network layer, the middleware layer, and the application layer [4,5]. The perception layer is used for identifying objects and transforming them into digital signals. The data are then collected by the network layer for further transmission to the upper layers. The connectivity of the IoT infrastructure is also overseen by the network layer. The processing of data and the connections to the cloud and the database are overseen by the middleware layer. The processed data from the middleware layer are received by the application layer. It is responsible for the implementation of applications under different use cases. The classification of various security attacks in IoT is given below and shown in Figure 2 [4]:

• Perception Layer: Unauthorized access to the tags [7], Tag cloning [8], Eavesdropping [9], RF jamming [10], Spoofing attack [11], Sleep deprivation attack [11].
• Network Layer: DoS attack [12], Sybil attack [13], Spoofed routing information attack [14], Selective forwarding attack [14], Sinkhole attack [15], Blackhole attack [15], Man-in-the-middle attack [16].
• Middleware Layer: Flooding attack [17], Cloud malware injection [18], Signature wrapping attack [19], SQL injection [19], Web browser attack in the cloud [17].
• Application Layer: Code injection [20], Buffer overflow [21], Sensitive data permission/manipulation [10], Phishing attack [22], Authentication and authorization [20].

To address these security issues, there are some common security requirements for IoT systems, as shown in Figure 3.

3. Proposed System

The proposed system is designed for applications in mission-critical or resource-constrained IoT applications, where security is paramount. The system aims to provide a comprehensive security solution that can be adapted to various IoT use cases by employing a combination of hardware-based protection, secure communication channels, and a custom authentication protocol. The key features of the system include a pressure-sealed casing for physical security, a one-way communication channel to prevent security attacks, and a tampering detection system that triggers an automatic reset and data wipe, if any unauthorized access is detected.

3.1. System Architecture

The proposed system architecture is built around several core components, shown in Figure 4. Each component contributes to the overall security of the IoT sensing device. These components include:

• Sensing Device: It is the central element of the IoT system, responsible for sensing and collecting data from its environment. It is connected to a processor which performs necessary operations on the data sensed by it.
• Memory: The memory stores the processor-executable instructions that dictate the system’s behavior. It also temporarily stores the sensor data before they are encrypted and transmitted.
• Custom Port and Master Device: The sensing device interfaces with a master device through a custom UART port. The custom port integrates an embedded authentication protocol, utilizing cryptographic handshakes to prevent unauthorized connections, ensuring high security in resource-constrained environments. The master device plays a crucial role in configuring the sensing device and sending the control signals. A 115,200 bps data rate is used, as it balances the need for real-time data transfer with minimal processing overhead, ensuring consistent performance in constrained environments
• Pressure-Sealed Casing: The IP65 casing houses the sensing device and is equipped with a pressure monitoring sensor to detect any unauthorized attempts to open it. A drop in the internal pressure indicates tampering, prompting the system to initiate security protocols.
• One-Way Communication Channel: To transmit data from the sensing device to a designated source address, the system utilizes a one-way communication channel and LoRa (Long Range) communication module. This allows only data transmission, and prevents any inbound communication attempts that could be exploited by attackers.
• Processor: It is the main control unit of the system and executes all the instructions stored in the memory for various tasks. It is also responsible for the authentication of the master device, the processing of the control signal, the encryption of data to be transmitted, and the monitoring of the pressure within the sealed casing to detect tampering attempts.

3.2. Methodology

To ensure the security of the sensing device and the integrity of the sensed data collected by it, the proposed systems use the following methodology, also depicted in Figure 5:

• Verification of the Master Device: The process begins with the docking of a master device to the custom port of the sensing device. The custom port ensures that only compatible devices can connect. Upon docking, the custom port authentication and handshake protocol is initiated, verifying the identity of the master device. The UART protocol is modified to include an additional control byte in each data packet, which serves as a unique signature to disambiguate the communication within the system. This step is crucial in preventing unauthorized devices from accessing the sensing device or from initiating man-in-the middle attacks.
• Reception of Control Signals: Once the master device is authenticated, the processor receives control signals from it. These signals may include configuration instructions or commands for the sensing device.
• Data Encryption and Transmission: The sensed data is processed by the processor and encrypted using the AES-128 encryption algorithm before the transmission, to prevent unauthorized access. This is achieved by a one-way communication channel, which restricts only the transmission of data and not its reception thereby reducing the risk of remote attacks.
• Tampering Detection: The system will continuously monitor the pressure within the sealed casing of the sensor system. If any pressure drop is detected, a tampering detection mechanism is activated, and a pre-defined series of steps, as a security measure, is initiated.
• Security Response: When tampering is detected, an alarm is triggered by the system to alert the user or system administrator. At the same time, the processor within the sensor system initiates a reset of the sensing device, thereby erasing all stored data and returning the device to its original factory settings. This ensures that sensitive data are always protected, even in the event of a physical security breach of the system.
• Post-event Handling: Once the device has been reset, it can only be reconfigured by an authorized master device, which again undergoes the previous steps of verification. This re-initialization process makes sure that the device is again secured and that no residual vulnerabilities from the previous tampered state are carried over.

The architecture and methodology proposed for the senor device are designed to address the security concerns of IoT sensing devices, specifically in resource-constrained applications. The system provides a solution for protecting the IoT devices by combining physical security measures with advanced authentication and secure communication protocols.

Table 1. Attack vectors and response of the proposed system.

Attack Vector	Description	System Response
Physical Tampering	Attempts to open or manipulate the device physically.	Device is disabled, all stored data is erased, and an alert is triggered.
DDoS Attacks	Distributed-denial-of-service attacks to overwhelm the device with network traffic.	Device remains unaffected, as the one-way communication channel does not allow inbound communication.
Device Signal Interference	Attempts to jam or interfere with the device’s wireless signals.	No impact, as the device does not use a wireless receiver for data reception, only one-way transmission.
Device Spoofing	An attacker attempts to mimic the device to impersonate its communication.	- Custom port does not allow docking/pairing - Even if the device docked, embedded digital signatures in communication packets ensure that only authenticated data is processed.

showcases the response of the proposed system against different attack vectors.

4. Discussion

An abstract example of an IoT device’s activity in relation to time is shown in Figure 6 [23], which also includes the four steps, their typical order, and some potential power consumption values. It is evident that the information that needs to be transferred may be temporarily kept in memory, before being sent. It is possible, of course, that data is delivered right away, following the processing. The memory capacity and energy usage for storing the data, as well as how these factors relate to the energy usage for activating and deactivating the wireless transmitter, will determine which strategy is best.

There are some factors that limit the power management of these IoT devices. These include voltage accuracy, sleep and shutdown cycles, efficiency of energy source, aging, scalability, etc. [24]. To further improve the energy efficiency of the IoT device, a power gating technique, as a power management solution, can be used to reduce the energy consumption during the sleep and active modes. One such technique is presented in [25].

Table 2. Power profile of a sensor node.

Operating Mode	MCU	Sensors	Radio
Active (regular operation)	5 mW	30 mW	350 mW
Inactive (power saving)	2 µW	5 µW	20 µW
Inactive (power gating)	2 µW	1 µW	2 µW

shows the performance improvement achieved using this technique. Two scenarios are compared: (a) the power-saving technique, that is the use of standby/sleep pins already available at the sensing and radio modules, and (b) the power-gating technique, implemented externally. A similar technique can be used in our proposed device in the future and its performance can be evaluated.

5. Conclusions

Owing to the large number of Internet of Things (IoT) devices in various applications, there is a need for an innovative solution to improve the security of the devices against the vulnerabilities in conventional IoT architectures, which can lead to high risks to data integrity, privacy, and system reliability. Considering this, in this paper, a novel approach towards securing IoT devices against any malicious attack is proposed. A new IoT node, housed in an IP65 casing, is proposed. It has a single communication channel, which sends information only to a specific endpoint and limits communication to a single dedicated device, thereby drastically reducing the attack probability. It can mitigate common attacks, such as eavesdropping, data tampering, and unauthorized access. The proposed system also has a triggering system to detect any physical or software tampering attempt. It also improves resource efficiency, lowers power consumption, and simplifies network management, thus making it suitable for resource-constrained applications, such as military surveillance, home automation, etc. In the future, research on the implementation of a power management approach, such as power gating, can be conducted, to further reduce the power consumption in these applications.

6. Patents

A patent for this work has been filed with an application number 202411044509 in the Indian Patent Office on 8 June 2024.