This article presents a detailed analysis of hardware encryptors and cryptographic libraries for ensuring security in IoT systems. The Internet of Things (IoT) is currently one of the most dynamic fields, where data security is becoming critically important. The study examines how various microcontrollers use hardware and software encryption methods to protect data and provides a comparative analysis of the effectiveness of these methods. Microcontrollers, which form the backbone of many IoT devices, perform tasks ranging from sensor reading to controlling various devices. To ensure the security of data transmitted through IoT systems, it is essential to use reliable encryption methods. Hardware encryptors embedded in microcontrollers provide high performance and energy efficiency. Software cryptographic libraries, such as mbed TLS, AESLib, TinyAES, and others, offer flexibility and can be used on various platforms. The article examines encryption methods, both hardware and software, using popular microcontrollers such as the STM32F407VG and ESP32-WROOM-32 as examples. Performance, energy consumption, and security levels were measured. Performance was assessed by determining the number of encryption operations per second, which allows evaluating the real-time encryption speed. Energy consumption was measured using a precision multimeter to determine the amount of energy consumed during encryption. Security level was assessed through an analysis of the physical security of keys and resistance to various types of attacks, including brute force and side-channel attacks. The results of the study showed that hardware encryption on the STM32F407VG microcontroller provides significantly higher performance, lower energy consumption, and higher security compared to software encryption on the ESP32-WROOM-32. This confirms the high efficiency of hardware encryption for use in IoT systems that require reliable data protection.
In today's technological landscape, the Internet of Things (IoT) is becoming an increasingly
significant part of our daily lives [
Data encryption is one of the most effective methods for protecting information in IoT. It allows
data to be transformed into a format that cannot be read without a special key, thus ensuring its
confidentiality even if intercepted [
Microcontrollers, which are the backbone of many IoT devices, use both hardware and software
encryption [
The purpose of this study is to evaluate the effectiveness of hardware and software encryption on microcontrollers in the context of IoT systems. The research aims to determine the performance, energy consumption, and security level of the two encryption approaches: hardware encryption on the STM32F407VG microcontroller and software encryption on the ESP32-WROOM32 microcontroller using the mbed TLS cryptographic library.
The analysis of existing works allows identifying the trends and challenges faced by IoT device
developers, as well as approaches that can be used to optimize security [
Several publications are dedicated to data protection tasks in IoT networks, with some studies
focusing on improving encryption methods [
Regarding the libraries for implementing these ciphers on microcontrollers, several researchers
[
Many studies also address the optimization of microcontroller resources when applying
encryption, particularly ways to reduce computational costs and memory requirements for
implementing encryption algorithms [
Despite some research in this area, there are certain unresolved issues and tasks that require further investigation. One of the main challenges is ensuring maximum data security with the limited resources of microcontrollers. Additionally, research on the effectiveness of encryption optimization methods to ensure low energy consumption and high data processing speed in IoT networks remains relevant.
Microcontrollers are the foundation of most IoT devices, providing data processing, communication, and device management. Their typically rich functional capabilities make them suitable for a variety of applications. Microcontrollers play a key role in IoT, as they are used to control and monitor connected devices and collect data from them.
The role of microcontrollers in an IoT system usually depends on their functionality and the tasks they are assigned. The use of microcontrollers in IoT has several typical directions, including:
Overall, microcontrollers are a fundamental component of IoT systems, as they perform critical tasks related to sensors, device control, communication, and data processing. In the context of these functions, ensuring data security becomes extremely important.
The requirements for protecting the confidentiality, integrity, and availability of data are the main reasons why encryption plays a crucial role in IoT. Microcontrollers collect and process a lot of sensitive information, and without proper protection, this data can become a target for cybercriminals. The use of modern cryptographic methods, such as hardware and software encryption, helps ensure reliable data protection at all levels of the IoT system.
Encryption on microcontrollers plays a critical role in ensuring data security in embedded systems,
particularly in the context of IoT [
Many modern microcontrollers are equipped with built-in hardware modules for encryption. These hardware encryptors allow cryptographic operations to be performed at the hardware level, ensuring high data processing speed and low power consumption. Hardware encryption significantly reduces the load on the microcontroller's central processor, freeing it to perform other important tasks.
For microcontrollers that do not have built-in hardware encryption modules, software cryptographic libraries are used. These libraries provide necessary functions for encryption, decryption, and key management, implementing cryptographic algorithms at the software level.
The features of hardware encryption and the use of cryptographic libraries will be discussed in the following sections, where we will analyze their advantages, disadvantages, and compare their effectiveness in various usage scenarios in IoT systems.
Hardware encryption is an essential component of the functionality of modern microcontrollers, providing high performance and low power consumption during cryptographic operations. Hardware encryptors are integrated directly into microcontrollers and allow encryption and decryption of data at the hardware level. Figure 1 illustrates the block diagram showing the main components of the hardware encryptor and their interaction.
From Figure 1, it can be seen that the input data is fed into the Encryption Processor, which performs the core cryptographic operations (Table 1) to transform Plaintext into Ciphertext.
Non-linear substitution of each byte of the plaintext using a substitution table (S-box) Cyclic shift of bytes in each row of the state matrix Linear transformation of each column of the state matrix Bitwise XOR between the state matrix and the round key ℎ
Formula ( ) = ( )
( ) = ( ) ( ) = ′ ℎ ( ) = ′′ ( ) = ′
( ) = ′′ ( , ) = ⊕ ( , ) = ⊕ The Key Generator block denotes the generation of cryptographic keys using a Pseudo-Random Number Generator (PRNG) to create reliable keys. These keys are stored in the key memory and passed to the encryption processor and control blocks. Keys are stored in The Key Storage and are only accessible to the encryption processor through the control blocks. Control Units manage the encryption process, provide operation synchronization, and control access to keys and data. Input/Output (I/O Interfaces) Interfaces facilitate data exchange between the hardware encryptor and other parts of the microcontroller or external devices. They transmit encrypted data to the next processing or storage stages. Encrypted data (Ciphertext) emerges at the output of the encryption processor, which can be safely transmitted or stored. Upon completion of the encryption process, the data will be protected from unauthorized access.
Let's consider the properties of popular microcontrollers and SoCs that have built-in hardware capabilities for encryption.
1. STM32. STM32 microcontrollers typically feature built-in AES (Advanced Encryption
Standard) hardware encryption, providing efficient and secure data encryption [
Hardware encryption provides significant advantages in performance and security, as encryption is performed by specialized components of the microcontroller. However, not all microcontrollers have built-in cryptographic accelerators. In such cases, software cryptographic libraries come to the rescue, implementing the necessary cryptographic algorithms and functions at the software level.
Cryptographic libraries play an important role in ensuring data security in microcontrollerbased systems, especially when hardware encryptors are unavailable or insufficient to perform all necessary cryptographic operations. These libraries implement various cryptographic algorithms and functions at the software level, providing flexibility and a wide range of capabilities for developers.
The first step in implementing encryption is initializing the necessary cryptographic libraries on the microcontroller. This stage involves selecting the cryptographic algorithm, generating keys, and configuring operating modes (encryption or decryption). Subsequent procedures involve data preparation, including dividing it into blocks if the algorithm requires working with data blocks, and adding additional information, such as an initialization vector, to enhance encryption security. Data encryption is performed according to the selected cryptographic algorithm and key, including adding data for authentication (if necessary), key expansion, and performing mathematical encryption operations on data blocks. Upon completion of the encryption operations, encrypted data is obtained, which can be used for transmission or storage in a secure location.
The list of popular cryptographic libraries relevant for microcontroller-based projects is quite extensive:
Crypto a cryptographic function library for Arduino and other microcontrollers [
TinyAES a library for implementing AES on Arduino microcontrollers. It is known for
its efficiency and small size, making it an ideal choice for resource-constrained embedded
systems. This library offers AES implementation on AVR microcontrollers (on which
Arduino Uno is based) and is optimized for the limited resources of these devices. It can be
particularly useful for projects where code size and speed are crucial [
uECC a library used to implement elliptic curve cryptography on microcontrollers [
Mbedtls a library offering implementations of a wide range of cryptographic algorithms,
but may be somewhat heavy for use on Arduino Uno platforms due to its large amount of
code and resource requirements [
Compatibility of encryption libraries with different microcontroller modifications is a crucial aspect in the development of security systems for embedded devices. The choice of an appropriate library depends on several factors, including the computational power of the microcontroller, the amount of available memory, energy consumption requirements, and support for specific cryptographic algorithms. Let's consider the main aspects of compatibility: 1. Computational Power. Microcontrollers with higher computational power (e.g., ARM Cortex-M series) can support more complex cryptographic algorithms and libraries, such as mbed TLS or WolfSSL. Less powerful microcontrollers (e.g., AVR or PIC) may require libraries optimized for them, such as TinyCrypt. 2. Memory. The amount of available RAM and Flash memory also affects the choice of library. Some encryption libraries require significant memory to store keys and perform algorithms, which can be an issue for microcontrollers with limited resources. For such cases, there are optimized libraries like micro-ecc for ECC cryptography. 3. Energy Consumption. For battery-powered devices, it is important to choose libraries optimized for low energy consumption. This might include the use of specific algorithms or hardware cryptography accelerators if they are available in the microcontroller. 4. Support for Cryptographic Algorithms. Different applications may require different cryptographic algorithms (AES, RSA, ECC, etc.). It is important to ensure that the library supports the necessary algorithms and that they are optimized for the specific microcontroller.
Comparison of the properties of cryptographic libraries for microcontrollers is presented in Table 2.
ESP8266, ESP32 STM32, AVR, ESP32 STM32, ESP32, ARM Cortex-M STM32, ESP32, NXP LPC
The study used two types of microcontrollers: one with hardware encryption (STM32F407VG) and one with software encryption (ESP32-WROOM-32). To evaluate the efficiency of hardware and software encryption, the performance, energy consumption and security metrics were used. The number of encryption operations per second was measured to assess how quickly encryption can be performed in real-time. Encryption performance was measured by determining the number of encryption operations executed per second. This allows for evaluating the real-time encryption speed and comparing the efficiency of different encryption approaches. For software encryption on the ESP32-WROOM-32, the mbed TLS cryptographic library was used. Each microcontroller was set up to perform encryption on a fixed-size data block (16 bytes) using the AES algorithm. Specific programs were developed to execute encryption in a loop for this purpose.
The testing procedure involved running the encryption program on the microcontroller and measuring the time required to perform a certain number of encryption operations (1000 operations). The number of encryption operations per second was then calculated by dividing the total number of operations by the total execution time. Data was collected for each type of encryption (hardware and software). The test results were recorded and analyzed to determine the average performance value, after which graphs were created to visualize the performance comparison between hardware and software encryption.
The measurements showed that hardware encryption on the STM32F407VG significantly outperformed software encryption on the ESP32-WROOM-32 in terms of the number of encryption operations per second (Figure 2). This confirms the high efficiency of hardware encryption for use in systems requiring fast and reliable data encryption.
Energy consumption was measured during the execution of encryption operations. This is crucial for devices with limited power resources, such as battery-powered devices. Software encryption on the ESP32-WROOM-32 was performed using the mbed TLS cryptographic library. Each microcontroller was configured to encrypt a fixed-size data block (e.g., 16 bytes) using the AES algorithm. Corresponding programs were developed to perform encryption in a loop.
Energy consumption was measured using a precision multimeter or similar equipment that can accurately measure the current and voltage consumed by the microcontroller. After starting the encryption program on the microcontroller, the current and voltage were measured during the encryption operations. Power was calculated, and the time required to perform a certain number of encryption operations (1000 operations) was measured. Data was collected for each type of encryption (hardware and software), and the measurement results were recorded and analyzed to determine the average energy consumption during the encryption operations. Graphs were then created to visualize the comparison of energy consumption between hardware and software encryption.
Measurements showed that hardware encryption on the STM32F407VG had significantly lower energy consumption compared to software encryption on the ESP32-WROOM-32 (Figure 2). This confirms the high efficiency of hardware encryption for use in energy-constrained systems that require effective encryption with minimal energy consumption.
The level of protection was assessed, including the physical protection of keys and resistance to attacks. Evaluating the security level of encryption involved analyzing the physical protection of keys and resistance to various types of attacks. This stage is critically important for ensuring reliable data protection, especially in the face of increasing threats in the field of information security.
Hardware encryption provides a higher level of security because keys are stored in specialized memory of the microcontroller, making unauthorized access more difficult. Software encryption was supported by the mbed TLS cryptographic library, which stores keys in the microcontroller's general memory, making them more vulnerable to attacks. Resistance to attacks was assessed by conducting simulations of various types of attacks, including brute-force attacks, side-channel attacks, and differential attacks. Tests were conducted for each microcontroller to model possible attack scenarios and determine how effectively each type of encryption could withstand these threats. Data on the physical protection of keys and the results of tests for resistance to attacks were collected and analyzed to determine the overall security level of each type of encryption. Graphs were then constructed to visualize the comparison of security levels between hardware and software encryption.
The results showed that hardware encryption on the STM32F407VG provides a significantly higher level of security compared to software encryption on the ESP32-WROOM-32 (Figure 2). This confirms the importance of using hardware encryption to protect critical data in the face of increasing threats in the field of information security.
Further research in the field of hardware and software encryption on microcontrollers could focus on several key aspects. Firstly, analyzing the efficiency of hardware encryption on new and emerging microcontrollers in the market would be worthwhile. This would allow assessing the potential of new hardware platforms in terms of security and performance.
Secondly, it's important to develop and optimize software cryptographic libraries to achieve better performance and energy efficiency across different hardware platforms. Optimization may include algorithm enhancements, reducing memory consumption, and lowering energy consumption, which is particularly crucial for resource-constrained devices.
Integrating hardware and software encryption with other security protocols, such as TLS, is another promising direction. This would provide comprehensive protection for IoT systems, enhancing their resilience to various types of attacks and ensuring secure data transmission.
Also, one of the directions of future research is the integration of post-quantum protocols in
encryption on microcontrollers for IoT devices [
Expanding the scope of testing is also an important aspect of further research. Conducting more extensive tests using different scenarios and data types would provide a more comprehensive picture of the efficiency and security of encryption under various conditions. This would help identify potential weaknesses and refine existing encryption methods.
Overall, further research has the potential to significantly enhance the security and efficiency of IoT systems. It would assist developers in selecting the best solutions for data protection, considering the specific requirements of their projects. The results of this research could serve as a basis for creating more reliable and energy-efficient IoT devices that meet modern security challenges.
Based on the research findings, it is recommended to use microcontrollers with built-in hardware encryption engines, such as STM32, for projects requiring high performance, low power consumption, and high security levels. Hardware encryption is the most efficient solution for critical IoT applications where data reliability and processing speed are crucial.
Software cryptographic libraries like mbed TLS, AESLib, TinyAES, and others remain essential tools for projects where hardware encryption engines are unavailable or where flexibility in choosing encryption algorithms is needed. The choice of a specific library should be based on the performance, power consumption, and security requirements of the particular project.
Thus, encryption on microcontrollers serves not only as a means of data protection but also as a critical component for supporting functional security and stability in IoT systems. Implementing effective encryption methods helps mitigate risks associated with data breaches, unauthorized access, and other cyber threats, ensuring the reliability and longevity of connected devices in the modern Internet of Things world.
erformance comparison of ECC libraries for IoT -Applied Sciences