Analysis of wireless transmission technology based on wearable medical system chips

　　Introduction

　　Health is closely related to everyone and has become a hot topic of concern in today's human society. Wearable medical monitoring systems can conveniently collect human health data for disease prediction and early diagnosis. The wearable medical chip system solution, based on low-cost, low-power, and high-transmission wireless communication technology, helps patients collect basic vital parameters in real time during daily work and life. By reducing face-to-face consultation time between doctors and patients, it shortens waiting times in hospitals, thereby alleviating the current shortage of medical resources and improving the quality of patient care. Additionally, chronic diseases (such as hypertension, diabetes, and high blood lipids) have become the number one killer of human health today. Treatment of chronic diseases requires long-term, continuous collection and monitoring of patients' health data. Wearable medical chips, due to their small size, low power consumption, and low operating costs, are easier for patients to accept. The vast potential consumer market has attracted many chip design companies such as Philips, Zarlink, Ti, etc. to participate in their R&D and commercial promotion.

　　Wrist-worn blood oxygen sensors, wristwatch-style blood glucose sensors, wristwatch-style sleep quality monitors, sleep physiology checkers, belt-style breathing and heartbeat monitors, implantable identity recognition components, and more. Wireless wearable medical microsystems consist of wireless sensors embedded on the body surface, such as everyday clothing, watches, jewelry, etc., all of which can be used to implant micro wearable medical chips. Because wireless communication technology is placed at different parts of the body surface, with numerous wires connecting different sensors and between the main processing display chips, it inevitably causes significant inconvenience to users. As an alternative transmission method for wires, wireless communication technology stands out as a particularly prominent advantage. Currently, most wireless communication technologies focus on increasing the transmission rate of wireless data, and wireless transmission technologies used in wearable medical systems must also consider minimizing power consumption during wireless signal transmission. The transceiver part used for wireless signal transmission on wearable medical chips is usually the most energy-consuming part of the entire medical chip. To facilitate long-term wearable use, the power consumption of the wireless transmission circuit is undoubtedly a key consideration for wearable chip designers. Focusing on the goals of low power consumption and high transmission rates, companies such as Zarlink, Nordic, Philips, and chipcon have successively launched solutions for ultra-low power RF transceiver chips.

　　1. Chip structure of wearable medical systems

　　The overall structure of a wearable medical chip based on wireless communication technology is shown in Figure 1, generally consisting of a physiological signal acquisition circuit, an analog-to-digital conversion circuit (ADC), a digital signal baseband processing circuit, a controller, and a power source

　　The receiver circuit consists of several parts. First, the signal acquisition low-noise instrument amplifier circuit collects physiological data from the human body. Then, the acquired physiological signals are converted through AD to quantify and generate easy-to-process digital signals. After encoding, FFT, and other digital signal processing, they are transmitted through the transmitting circuit. At the same time, external control signals and data can also be received through the receiving circuit on the chip. The controller is used to control the operation of the entire chip and can be programmed to meet different application requirements. Typically, a high-performance wearable medical chip consists of high-performance digital, analog, and RF components, with the performance of these components in particular directly affecting the overall chip performance. The analog and RF transceiver parts of medical chips are obviously the most power-consuming parts of the entire chip, so designers usually have to balance low power consumption and high performance when designing circuits for these two parts. Below, we introduce the various components of a typical wearable medical system chip.

　　Figure 1: Structure diagram of the wearable medical chip system

　　1.1. Physiological signal acquisition low-noise amplifier

　　Physiological signals are generally collected through on-chip integrated biosensors. To facilitate integration, the sensor uses a low-noise amplifier using CMOS process to convert biological signals into bioelectrical signals. To simultaneously obtain multiple physiological information, multiple amplifiers with different functions can be integrated on the chip to form multi-channel data to collect vital parameters such as blood pressure, blood oxygen saturation, respiratory rate, heart rate, and body temperature. Since physiological signals in the human body are relatively weak and easily affected by surrounding noise, amplifiers must achieve high sensitivity, high gain, low noise, and low power consumption; At the same time, a low-pass filter with a cutoff frequency of around 1kHz is used after the amplifier to further filter out interference noise at frequencies other than bioelectric signals. Amplifiers can be designed with multiple operating modes, such as listening, working, and sleeping, to reduce chip power consumption.

　　1.2. AD Converter (ADC)

　　The pre-mounted multi-channel physiological signal acquisition amplifier collects various physiological information and connects to the ADC's input port via an analog multi-channel multi-connector. The analog multi-channel multi-channel amplifier can only select the output of one preamplifier at a time. To reduce power consumption, ADCs typically use a sequential approximation structure, with about 10 bits. To improve accuracy and conversion speed, sigma-delta or pipeline ADCs can also be used. The higher the bit count, the higher the conversion rate, but the higher the power consumption. Low power consumption is key when designing wearable medical chips. Additionally, the ADC's unit capacitance should be chosen appropriately; selecting too large takes up a lot of chip space, and the impact of parasitic capacitance on the unit capacitance should be minimized.

　　1.3 Controller

　　The chip can use ARM cores and MCU as controllers, controlling the operating modes of other parts of the chip's circuits via bus; It can control the timing of data usage, configure registers, and control other parts of the chip to occupy the data bus for real-time communication.

　　1.4 Digital signal processing baseband

　　To improve the speed, accuracy, and security of data transmission, the digital signals output by the ADC must go through the baseband processor of the digital signal controller, undergo digital compression and encoding, and can also be further filtered out through FFT transformation and digital filtering to further filter out interference frequency noise.

　　1.5, RF transceivers

　　Since the collection of physiological signals from the human body requires physiological characteristics, placing wearable medical chips in different parts of the body, and the presence of interconnected wires between chips makes movement inconvenient, and too many wires can easily get tangled and cause great discomfort. Therefore, wireless transmission of signals and data is the most direct and natural method. The key issues to consider when integrating wireless RF transceivers on wearable medical system chips differ significantly from those typically addressed in wireless product applications. First, this is an asymmetric wireless transmission method, mainly collecting human signals and sending them out. The received signals mainly come from control commands, and the data volume is very small. Therefore, a half-duplex communication mode can be used, with low-speed downlink and high-speed uplink transmission. Secondly, chips need to operate for long periods, and the batteries used for wearable chips are generally button cells, operating at voltages between 1.2~1.5 V and with capacity less than several hundred mA·h. The wireless transceiver section is usually the part with the highest power consumption in a chip. Designers face challenges such as low operating voltage, low power consumption, and high transmission rates. Therefore, it is necessary to carefully consider the structure adopted by the wireless transceiver, as well as the implementation of key technologies such as carrier frequency, transmission method, modulation method, transmission rate, and power consumption.

　　2. Wireless communication standards for wearable medical chips

　　Wireless communication technology is advancing rapidly, playing a significant role in advancing modern medical technology. Currently, there are various communication standards available for communication between wearable medical chips. These standards are suitable for specific applications based on their own characteristics, but may also fail to fully utilize the low-power, short-range communication features of wearable medical chips. Below is a brief introduction to the performance and characteristics of each communication standard (see Figure 2).

　　Figure 2: Comparison of transmission distances and power consumption of various wireless communication methods

　　2.1 Bluetooth

　　The Bluetooth standard uses frequency hopping and spread spectrum technology, which effectively suppresses inter-code interference, improves communication quality, and maintains call security. Bluetooth standards support three different communication distances: 1, 10, and 100 m, and can provide communication speeds up to 1 Mbps. It has a simple structure and can reduce the price of a single chip to below $5, with mature technology and strong market competitiveness. The Bluetooth standard provides point-to-point serial communication and a shared channel main controller interface communication method, which is very suitable for building human local area networks. However, since the communication range of wearable medical chips is generally limited to areas close to the human body, while Bluetooth operates at 2.4 GHz, the impact of such high frequencies on the human body remains unknown. Due to people's fear of high-frequency communication and its relatively high power consumption, the Bluetooth standard is not an ideal choice.

　　2.2，Zigbee

　　Zigbee can operate at three different frequency ranges: 2.4 GHz, 900 MHz, and 800 MHz. Compared to Bluetooth standards, Zigbee consumes less power. When operating in the 2.4 GHz band, it can reach a maximum data transfer rate of 240 kbps. Zigbee's drawbacks are low data transmission rates, high transmission latency, poor security, and when operating at the 2.4 GHz frequency, the wide variety of communication protocols concentrated in that frequency band makes Zigbee easily susceptible to interference from other communication waves.

　　2.3，UWB

　　UWB operates in the frequency range of 3.1~10 GHz, with an average data transmission rate of up to 850 kbps and can be increased to 26 Mbps. This standard specifies a power spectral density of -41dB(m) MHz, but there are no specific requirements for time-domain waveforms. Therefore, pulse transmission technology can be used, making the structure of RF transmitters very simple, while transferring design pressure and power consumption design to RF receiver design. As mentioned earlier, wearable medical chips transmit asymmetric signals, with transmitted data flow far exceeding input data flow, making UWB well-suited to this asymmetric wireless communication feature, thereby reducing power consumption and system complexity. Moreover, UWB is an ultra-wideband technology that uses ultra-wide band to achieve lower power consumption, resulting in relatively low power consumption.

　　2.4，WLAN 802.11

　　IEEE 802.11 WLAN operates in the ISM band (industrial, scientific, and medical bands). Among them, 802.11b and 802.11g operate in the 2.4 GHz band, with data transfer rates of 11 Mbps and 54 Mbps, respectively. The 802.11a operates on the 5 GHz band and can provide transmission rates of up to 54Mbps. It has a relatively long communication range and, due to its use of direct sequence spread spectrum technology, has strong anti-interference capability. However, it consumes a lot of power, has a complex structure, and is too expensive, making it unsuitable for the design of wearable medical chips.

　　2.5, Wireless USB

　　Wireless USB technology, like UWB, is a wireless communication technology based on ultra-wideband technology. It operates in the 3.1~10.6 GHz range, with communication distances of 3 and 10 m, suitable for short-range wireless data transmission, with data transmission rates up to 480 Mbps and 110 Mbps respectively. However, the biggest challenge facing this technology is power consumption, which is also the biggest limiting factor for its application in medical chip communications.

　　2.6. Infrared Communication (IrDa)

　　Infrared communication is a low-cost and simple wireless communication method, but due to the direct emission nature of infrared, IrDA is only suitable for Huang Jin and others in phase 5834: wireless transceivers based on wearable medical system chips have short distances, point-to-point alignment, and low transmission speeds. Compared to wireless communication technologies like Bluetooth and Zigbee, it is extremely inconvenient to use.

　　2.7. Radio Frequency Identification Technology

　　RFID technology is a type of RFID technology that uses space-coupled alternating electromagnetic fields to achieve data communication without human contact. China's planned RFID frequency band is 50~190 kHz, the high-frequency band is 13.56 MHz± 7 kHz, and there is also 432~434.79 MHz; Another planned frequency band in China is 900, 910, and 910.1 MHz, which have been widely used for train vehicle identification. Like IrDa and Zigbee, RFID is an indoor wireless communication technology with a short communication distance, making it useful in various medical applications such as mobile asset management, inventory management, real-time patient monitoring, drug tracking, and distribution. However, this technology itself is an electronic tag and RFID technology, with extremely low transmission rates and information that is easily stolen, making it unsuitable for real-time wireless connectivity applications in wearable medical chips.

　　2.8. Human communication

　　Human communication (Bio-channel) technology, also known as human communication technology, is a new concept that has emerged in recent years. It was first proposed by Zimmerman of MIT's Media Lab in 1995. Unlike any previous wireless communication technology, human communication uses the proximity of the human magnetic field or the human body itself as the communication medium. The communication distance is very short, sometimes requiring human contact to communicate. Therefore, it allows precise control of the communication range and the communication target, greatly reducing interference between different channel signals and ensuring communication security. Typically, communication in areas close to the human body can also be wired, which ensures high-speed and accurate data transmission without interference from external noise. However, wires tend to get tangled and are extremely inconvenient for people to use. On the other hand, using mature data communication technologies like Zigbee and Bluetooth avoids the hassle caused by wires, but also faces issues such as slow communication speeds, high chip power consumption, and susceptibility to interference from space clutter electromagnetic signals. Therefore, as soon as the concept of human communication was proposed, it immediately attracted widespread attention from academia and industry.

　　3. Development example of wireless transceivers based on wearable medical system chips

　　Due to the rapid development of microelectronics technology and the needs of an aging human society, wearable medical monitoring systems have been developed. A Body Area Network (BAN) consists of many human sensor nodes, each of which can communicate with other nodes (or central nodes) through wireless transceivers within the wearable medical chip. Early short-range wireless communication chip research for human medical monitoring often used ASK FSK modulation, low power consumption, and simple crystal oscillators as transmitters. This structure could only transmit single body sign data, had low performance, and had low oscillator frequencies and long switching and startup times, resulting in very low communication transmission rates. With the deepening of modern biomedical engineering research, over the past decade or so, some new circuits and systems based on inductive coupling coil communication have been proposed. However, these inductive coil-based solutions also suffer from poor communication quality, low transmission rates, and long transmission times, which effectively reduce communication efficiency and effectively shorten the battery usage time.

　　These non-standardized communication systems struggle to meet the demands for ultra-low power consumption, ultra-small size, high reliability, and high communication speed for wearable medical wireless communication. Driven by the growing demand for wireless health monitoring, research institutions and major chip companies worldwide have competed to conduct extensive application research and development in this field. Among the most representative examples are: Zarlink in Canada, which developed ZL70101 RF transceiver chips, The Sensium system-on-chip developed by Toumaz in the UK, as well as a low-power RF transceiver with a 2.4 GHz 400mV power supply voltage designed by the Wireless Node Network Communication Chip Research Group at UC Berkeley University in the United States, and a human communication wireless transceiver chip developed by the Korean Academy of Science.

　　3.1 Zarlink Medical Implantable Communication System ZL70101 Chip

　　In 2006, Canada's Zarlink Semiconductor Company launched an ultra-low power high-performance RF transceiver ZL70101 for medical implant systems. This chip is highly integrated; excluding network matching, it only requires one 24 MHz quartz crystal and two decoupling capacitors, totaling three off-chip components; Its operating frequency band is the 433 MHz ISM band, using a 0.18μm RF CMOS process. The transceiver operates at 5.5 mA, and in sleep mode, it is only 250 nA. The entire chip integrates a 400 MHz RF transceiver, a 2.45 GHz wake-up signal monitoring receiver, and one media path controller (MAC). The chip structure diagram is shown in Figure 3.

　　The receiver adopts a low-intermediate frequency structure, consisting of a low-noise amplifier, a mirrored frequency suppression mixer, an IFF multiphase filter (PPF), a signal strength indicator (RSSI), and an ADC. The transmitter consists of an upper mixer and a power amplifier, using FSK frequency shift keying modulation method. The wake-up system is a receiver using OOK modulation and operates in the 2.45 GHz band. It can periodically detect start signals from base stations to power on the entire chip, greatly reducing the average operating current of the chip. This chip is designed for implantable medical monitoring applications, but thanks to its ultra-low power design, 2m communication distance, and transmission rate up to 800kbps, it also excels at meeting the wireless connectivity requirements of external wearable medical chips.

　　Figure 3 Block diagram of Zarlink's MICS RF transceiver principle

　　3.2 Toumaz wireless transceiver for ultra-low power system chips for biological remote sensing

　　In 2007, Toumaz in the UK released a system integration chip called Sensium, which combined the SPI bus, ADC, MCU, SRAM, and an ultra-low-power RF transceiver. The RF transceiver section of this Sensium chip has a chip area of 7 mm², uses 0.13μm RF CMOS process, operates at 1 V, and operates at both the European standard 870 MHz band and the US standard 928 MHz band. The current consumption during reception is only 2.1 mA, the transmit power is -7 dB(m), and the transmit current is 2.6 mA; The transmit/receive section operates in a half-duplex mode, FSK modulation, with a bit error rate of 10-3 and a data transmission rate of 50 kbps. Since this chip was developed for telemetry and acquisition applications such as ECG, Xinbo, and body temperature, its performance indicators fully meet the requirements for design applications. The chip adopts a Sliding-IF structure, which offers higher image frequency suppression compared to traditional low-IF transceivers, and because it uses two-stage frequency migration, it has much less DC drift than zero-IF transceivers.

　　To meet the requirements for low power consumption, the entire chip operates at 1 V, which is less than the sum of Vth of PMOS and NMOS under 0.13μm processes. Therefore, many devices, especially those in analog and RF sections, operate in subthreshold and weak reflective regions, greatly reducing power consumption but also posing challenges for RF analog circuit design. The receiving section adopts a zero-IF structure, and the system structure of the entire chip is shown in Figure 4.

　　The LNA uses a single-ended input common-source and gate structure, with output using on-chip planar inductors and adjustable capacitance matrices as matching loads. The LNA output is directly connected to one end of the first-stage lower mixer, while the other input of this double-balanced Gilbert unit mixer is connected to the power supply, forming a pseudo-differential operating mode mixer structure. The final stage of the drive buffer in the transmitting section uses a single-transistor NMOS amplifier with an open-drain structure, with its drain directly connected to the off-chip inductor-capacitance matching network. The drain stage of this NMOS transistor is directly connected to the power supply, so a thick-gate double-gate NMOS transistor must be used to prevent chip breakdown. The transmitter has a simple structure, and its VCO operates in a self-oscillating state. Communication loss in the communication link can be adjusted through RSSI-based automatic gain control (AGC), and the gain of the transmitter's drive buffer can be adjusted, thereby improving power transmission efficiency.

　　3.3 Wireless transceiver chip based on human body communication

　　In 2007, a research team led by Seong-Jun Son at the Korean Academy of Sciences designed the world's lowest power consumption and a bio-channel wireless transceiver chip capable of transmitting data at 2 Mbps [55]. This chip uses broadband communication technology similar to UWB, relying on the body's near-magnetic field to transmit communication data. The entire transceiver integrates a fully digital transceiver system (see Figure 5), with no digital modulation. The chip operates at 1 V, with a power consumption of only 0.2 mW and a chip area of 0.85 mm². Its overall performance makes it highly suitable for interconnecting wearable chips requiring short distances, high data transfer rates, and extremely low power consumption.

　　Because this chip is designed based on human communication principles, its operating frequency can be 1~200 MHz, using a 0.25μm CMOS process. The entire transceiver chip has only one signal conduction electrode that contacts human skin or attaches to clothing, eliminating the need for additional global grounding electrodes required for traditional wireless human communication. The transmitter section of the chip mainly consists of a ring oscillator, pseudo-random code generator (PRBS), and driver buffer. The receiver section of the chip consists of an analog front-end amplifier, level-shifting circuit, Schmitt trigger, and clock recovery phase-locked loop (CDR) circuit. To reduce power consumption, the chip adopts modulation-free direct digital transmission, employing 200 MHz broadband data transmission, a fully digital clock recovery circuit, a fully numerically controlled digital oscillator (DCO), and quadrature sampling technology. The application of these low-power circuit design technologies minimizes power consumption in the most power-consuming front-end amplifier circuits and clock generation circuits.

　　Figure 4 Block diagram of Toumaz company's RF transceiver principle

　　Figure 5: Wireless transceiver based on human communication principles

　　4. Outlook for Wearable Medical Wireless Transceiver Chips

　　In today's society, people face tremendous pressure from work and life. As people's demands for their health continue to rise, wearable medical chips are gradually being integrated into everyday life. With the continuous development of biomedical engineering and microelectronics technology, wearable medical chips are gradually becoming more miniaturized and networked. Wearable medical microsystems require physiological signal sensor nodes to be worn on the patient, so miniaturization is needed to keep patients under low load during long-term wear. At the same time, physiological characteristic signals of patients need to be transmitted via wireless networks to central base station nodes or other sensor nodes, making networking the most fundamental requirement for development. Therefore, current wearable medical chips are inevitably moving toward fully integrated SoCs to achieve miniaturization and low cost; At the same time, the on-chip integrated RF transceiver circuit also allows sensor node signals to be conveniently and real-time transmitted, enabling mobile monitoring of human health status anytime and anywhere.

　　Currently, there is no dedicated wireless communication standard for personal wearable medical systems internationally. The IEEE802.15 series of standards, targeting industrial, home, and medical low-cost, low-power wireless communication markets, is used for the development of personal wearable medical chips. Although wearable medical chips based on Zigbee, Bluetooth, and WLAN have already been developed, their communication protocols are not specifically designed for wearable medical applications. Their MAC layer and QoS cannot be optimized for the low power consumption, high transmission speed, and short distance characteristics of wireless medical data transmission, so they do not yet meet application requirements. Faced with these challenges, medical chip designers still have significant room to develop in low-power circuit design and wireless communication transmission methods. Many innovative circuit system structures and concepts based on these considerations still require further research and improvement in practicality. With the development of wireless communication technology, improvements in integrated circuit technology, and continuous development of application markets, these issues will inevitably be resolved and lead modern human healthcare projects toward low-cost, miniaturized, intelligent, and networked development.