Wireless networks of electronically controlled implantable medical sensors and actuators will be the basis of many innovative and potentially revolutionary therapies. The biggest obstacle in realizing this vision of networked implants is posed by the dielectric nature of the human body, which strongly attenuates radio-frequency (RF) electromagnetic waves. In this paper we present the first hardware and software architecture of an Internet of Medical Things (IoMT) platform with ultrasonic connectivity for intra-body communications. This platform can be used as a basis for building future IoT-ready medical implants and wearable devices. We demonstrate that ultrasonic waves can be efficiently generated and received with low-power mm-sized components, and that despite the conversion loss introduced by ultrasonic transducers, the attenuation of 2.4 GHz RF is substantially greater than ultrasound, e.g., ultrasound attenuates 70 dB less than RF at distances over 10 cm. We show that the proposed IoMT platform requires significantly less transmission power than 2.4 GHz RF with equal reliability, e.g., 35 dBm less at distances over 12 cm with 10-3Bit Error Rate (BER) thus enabling a lower energy per bit and a longer device lifetime. Finally, we experimentally establish 2.4 GHz RF links do not function at distances greater than 12 cm, while ultrasonic links achieve a reliability of 10-6BER up to a distance of 20 cm with less than 0 dBm transmission power.

In Section VI of the above paper[1], incorrectly numbered reference citations were introduced during the editing of the paper. The citations should be as follows.

Wirelessly networked systems of implantable medical devices endowed with sensors and actuators will be the basis of many innovative, sometimes revolutionary therapies. The biggest obstacle in realizing this vision of networked implantable devices is posed by the dielectric nature of the human body, which strongly attenuates radio-frequency (RF) electromagnetic waves. In this paper we present the first hardware and software architecture of an Internet of Medical Things (IoMT) platform with ultrasonic connectivity for intra-body communications that can be used as a basis for building future IoT-ready medical implantable and wearable devices. We show that ultrasonic waves can be efficiently generated and received with low-power and mm-sized components, and that despite the conversion loss introduced by ultrasonic transducers the gap in attenuation between 2.4 GHz RF and ultrasonic waves is still substantial, e.g., ultrasounds offer 70 dB less attenuation over 10 cm. We show that the proposed IoMT platform requires much lower transmission power compared to 2.4 GHz RF with equal reliability in tissues, e.g., 35 dBm lower over 12 cm for 10-3Bit Error Rate (BEr) leading to lower energy per bit and longer device lifetime. Finally, we show experimentally that 2.4 GHz RF links are not functional at all above 12 cm, while ultrasonic links achieve a reliability of 10-6up to 20 cm with less than 0 dBm transmission power.

Biomedical systems of miniaturized implantable sensors and actuators interconnected in an intra-body area network could enable revolutionary clinical applications. Given the well-understood limitations of radio frequency (RF) propagation in the human body, in our previous work we investigated the use of ultrasonic waves as an alternative physical carrier of information, and proposed Ultrasonic WideBand (UsWB), an ultrasonic multipath-resilient integrated physical and medium access control (MAC) layer protocol. In this paper, we discuss the design and implementation of a software-defined testbed architecture for ultrasonic intra-body area networks, and propose the first experimental demonstration of the feasibility of ultrasonic communications in tissue mimicking materials. We first discuss in detail our FPGA-based prototype implementation of UsWB. We then demonstrate how the prototype can flexibly trade performance off for power consumption, and achieve, for bit error rates (BER) no higher than 10-6, either (i) high-data rate transmissions up to 700 kbit/s at a transmit power of -14 dBm (≈ 40 μW), or (ii) low-data rate and lower-power transmissions down to -21 dBm (≈ 8 μW) at 70 kbit/s. We demonstrate that the UsWB MAC protocol allows multiple transmitter-receiver pairs to coexist and dynamically adapt the transmission rate according to channel and interference conditions to maximize throughput while satisfying predefined reliability constraints. We also show how UsWB can be used to enable a video monitoring medical application for implantable devices. Finally, we propose (and validate through experiments) a statistical model of small-scale fading for the ultrasonic intra-body channel.

Wearable medical devices with wireless capabilities have become the cornerstone of many revolutionary digital health applications that promise to predict and treat major diseases by acquiring and processing physiological information. Existing wireless wearable devices are connected through radio frequency electromagnetic wave carriers based on standards, such as Bluetooth or Wi-Fi. However, these solutions tend to almost blindly scale down traditional wireless technologies to the body environment, with little or no attention to the peculiar characteristics of the human body and the severe privacy and security requirements of patients. We contend that this is not the only possible approach, and we introduce U-Wear, the first networking framework for wearable medical devices based on ultrasonic communications. U-Wear encloses a set of physical, data link, and network layer functionalities that can flexibly adapt to application and system requirements to efficiently distribute information between ultrasonic wearable devices. U-Wear also offers reconfiguration functionalities at the application layer to provide a flexible platform to develop medical applications. We design two prototypes that implement U-Wear and operate in the near-ultrasonic frequency range using commercial-off-the-shelf (COTS) speakers and microphones. Despite the limited bandwidth, i.e., about 2 kHz, and COTS hardware components not optimized for operating at high frequency, our prototypes: 1) achieve data rates up to 2.76 kbit/s with bit-error-rate lower than 10-5using a transmission power of 13 dBm (20 mW ); 2) enable multiple nodes to share the medium; and 3) implement reconfigurable processing to extract medical parameters from sensors with high accuracy.

It is well known that electromagnetic radio-frequency (RF) waves that are the basis of most commercial wireless technologies are largely unsuitable to interconnect deeply implanted medical devices. RF waves are in fact absorbed by aqueous biological tissues and prone to malicious jamming attacks or to environmental interference from pervaslvely deployed RF communication systems; moreover, they pose a potential safety hazard when exposure of tissues is prolonged and at high power. While existing wireless technologies can satisfy the requirements of some specific applications, the root challenge of enabling networked intra-body miniaturized sensors and actuators that communicate through body tissues is largely unaddressed. Considering these limitations, this article proposes a high data rate ultrasonic communication scheme for wireless intra-body networks. The proposed scheme can enable various applications that require high sampling rates such as neural data recording or monitoring of the digestive tract through endoscopic pills. The proposed scheme is based on Orthogonal Frequency-Division Multiplexing (OFDM), which is proven to be robust against frequency-selective channels with relatively long delay spreads like the intra-body ultrasonic channel. The proposed scheme is implemented in a prototype ultrasonic software-radio and demonstrated to achieve data rates up to 28.12 Mbit/s through synthetic phantoms mimicking the ultrasonic propagation characteristics of biological tissues.

Most research in body area networks to date has focused on traditional RF wireless communications, typically along the body surface. However, the core challenge of enabling networked intra-body communications through body tissues is substantially unaddressed. RF waves are in fact known to suffer from high absorption and to potentially lead to overheating of human tissues. In this paper, we consider the problem of designing optimal network control algorithms for distributed networked systems of implantable medical devices wirelessly interconnected by means of ultrasonic waves, which are known to propagate better than radio-frequency electromagnetic waves in aqueous media such as human tissues. Specifically, we propose lightweight, asynchronous, and distributed algorithms for joint rate control and stochastic channel access designed to maximize the throughput of ultrasonic intra-body area networks under energy constraints. We first develop (and validate through testbed experiments) a statistical model of the ultrasonic channel and of the spatial and temporal variability of ultrasonic interference. Compared to in-air radio frequency (RF), human tissues are characterized by a much lower propagation speed, which further causes unaligned interference at the receiver. It is therefore inefficient to perform adaptation based on instantaneous channel state information (CSI). Based on this model, we formulate the problem of maximizing the network throughput by jointly controlling the transmission rate and the channel access probability over a finite time horizon based only on a statistical characterization of interference. We then propose a fully distributed solution algorithm, and through both simulation and testbed results, we show that the algorithm achieves considerable throughput gains compared with traditional algorithms.

We consider the problem of designing optimal network control algorithms for distributed networked systems of implantable medical devices wirelessly interconnected by means of ultrasonic waves, which are known to propagate better than radio-frequency electromagnetic waves in aqueous media such as human tissues. Specifically, we propose lightweight, asynchronous, and distributed algorithms for joint rate control and stochastic channel access designed to maximize the throughput of ultrasonic intra-body area networks under energy constraints. We first develop (and validate through testbed experiments) a statistical model of the ultrasonic channel and of the spatial and temporal variability of ultrasonic interference. Compared to in-air radio frequency (RF), human tissues show a much lower propagation speed, which further causes unaligned interference at the receiver. It is therefore inefficient to perform adaptation based on instantaneous channel state information (CSI). Based on this model, we formulate the problem of maximizing the network throughput by jointly controlling the transmission rate and the channel access probability over a finite time horizon based only on a statistical characterization of interference. We then propose a fully distributed solution algorithm, and through both simulation and testbed results, we show that the algorithm achieves considerable throughput gains compared with traditional algorithms.

Wearable medical sensing devices with wireless capabilities have become the cornerstone of many revolutionary digital health applications that promise to predict and treat major diseases by acquiring and processing health information. Existing wireless wearable devices are connected through radio frequency (RF) electromagnetic wave carriers based on standards such as Bluetooth or WiFi. However, these solutions tend to almost-blindly scale down traditional wireless technologies to the body environment, with little or no attention to the peculiar characteristics of the human body and the severe privacy and security requirements of patients. We contend that this is not the only possible approach, and we present U-Wear, the first networking framework for wearable medical devices based on ultrasonic communications.

The use of wirelessly internetworked miniaturized biomedical devices is promising a significant leap forward in medical treatment of many pervasive diseases. Recognizing the limitations of traditional radio-frequency wireless communications in interconnecting devices within the human body, in this paper, we propose for the first time to develop network protocols for implantable devices based on ultrasonic transmissions. We start off by assessing the theoretical feasibility of using ultrasonic waves in human tissues and by deriving an accurate channel model for ultrasonic intrabody communications. Then, we propose a new ultrasonic transmission and multiple access technique, which we refer to as Ultrasonic WideBand (UsWB). UsWB is based on the idea of transmitting information bits spread over very short pulses following a time-hopping pattern. The short impulse duration results in limited reflection and scattering effects, and the low duty cycle reduces the impact of thermal and mechanical effects, which may be detrimental for human health. We then develop a multiple access technique with distributed control to enable efficient simultaneous access by mutually interfering devices based on minimal and localized information exchange and on measurements at the receiver only. Finally, we demonstrate the performance of UsWB through a multiscale simulator that models the proposed communication system at the acoustic wave level, at the physical (bit) level, and at the network (packet) level. We also validate the simulation results by comparing them to experimental results obtained with a software-defined testbed.

Biomedical systems of implanted miniaturized sensors and actuators interconnected into an intra-body area network could enable revolutionary healthcare and clinical applications. Given the well-understood limitations of radio frequency (RF) propagation in the human body, in our previous work we investigated the use of ultrasonic waves as an alternative physical carrier of information [1], and proposed Ultrasonic WideBand (UsWB), an ultrasonic multipath-resilient integrated physical and medium access control (MAC) layer protocol [2]. In this paper, we discuss the design and implementation of a software-defined testbed architecture for ultrasonic intra-body area networks, and propose the first experimental demonstration of the feasibility of ultrasonic communications in biological tissues. We first discuss in detail our FPGA-based prototype implementation of UsWB. We then demonstrate how the prototype can flexibly trade performance off for power consumption, and achieve, for bit error rates (BER) no higher than 10-6, either (i) high-data rate transmissions up to 700 kbit/s at a transmit power of -14 dBm (≈ 40 μW), or (ii) low-data rate and lower-power transmissions down to -21dBm (≈ 8μW) at 70kbit/s. Finally, we show how the UsWB MAC protocol allows multiple transmitter-receiver pairs to coexist and dynamically adapt the transmission rate according to channel and interference conditions to maximize throughput while satisfying predefined reliability constraints.

The use of miniaturized biomedical devices implanted in the human body and wirelessly internetworked is promising a significant leap forward in medical treatment of many pervasive diseases. Recognizing the well-understood limitations of traditional radio-frequency wireless communications in interconnecting devices within the human body, in this paper we propose to develop network protocols for implantable devices based on ultrasonic transmissions. We start off by assessing the feasibility of using ultrasonic propagation in human body tissues and by deriving an accurate channel model for ultrasonic intra-body communications. Then, we propose a new ultrasonic transmission and multiple access technique, which we refer to as Ultrasonic WideBand (UsWB). UsWB is based on the idea of transmitting information bits spread over very short pulses following a time-hopping pattern. The short impulse duration results in limited reflection and scattering effects, and its low duty cycle reduces the thermal and mechanical effects, which are detrimental for human health. We then develop a multiple access technique with distributed control to enable efficient simultaneous access by interfering devices based on minimal and localized information exchange and on measurements at the receiver only. Finally, we demonstrate the performance of UsWB through a multi-scale simulator that models the proposed communication system at the acoustic wave level, at the physical (bit) level, and at the network (packet) level.

Nanomachine-to-neuron communications are envisaged as a futuristic communication scenario with enormous impact especially on medical applications. One of the crucial aspects to be considered in the study of any communication environment is represented by the amount of information that can be reliably transmitted over the channel, that is, its capacity. In this paper we evaluate such capacity. To this purpose we consider that the communication channel between a nanomachine and a neuron has been proved to introduce a noise. This however cannot be modeled by means of a traditional additive white Gaussian noise as assumed in the traditional Shannon communication theory. Capacity and bit error rate are useful metrics to characterize the theoretical bounds for communications and the maximum achievable throughput.