Technical requirements for wireless body area networks (WBANs)

Technical requirements for wireless body area networks (WBANs)#

Based on the paper Technological Requirements and Challenges in Wireless Body Area Networks for Health Monitoring: A Comprehensive Survey by Zhong et.al..

Introduction#

demographic change due to population ageing. The proportion of the elderly increases.

  • elderly are more prone to suffer from chronic diseases

    • diabetes

    • being overweight

    • Alzheimer’s, Parkinson’s

  • these diseases are caused by

    • sedentary behavior

    • inappropriate eating habits

  • these behavior can be prevented by real-time monitoring of human body

WBAN

  • connects people and things in typical ~2m range

  • collected example data

    • electrocardiogram (ECG)

    • electroencephalogram (EEG)

    • blood oxygen saturation

    • blood pressure

    • heart rate variability (HRV)

  • terminal is a gateway to other networks like internet

  • WBAN can both sense and provide feedback to the body

  • WBANs are still in their infancy

  • combination of multiple disciplines (communication, bio-engineering and microelectronics) makes it difficult to solve key issues

Figure 1

!define gateway internet gateway <:globe_with_meridians:> !define cloud cloud <:cloud:> !define hospital hospital <:hospital:> !define wban WBAN nodes <:standing_person:> !define terminal terminal <:satellite:> (wban) -- (terminal): short range, Bluetooth, .. (terminal) -- (gateway): LAN, WLAN, Bluetooth ... (gateway) -- (cloud): wireless broadband (cloud) -- (hospital): broadband (hospital) -- (family <:family:>) (hospital) -- (mobile clinic <:minibus:>) (hospital) -- (remote diagnosis, etc.)

Three tiers

  • intra-WBAN

    • sensors preserve star topology

    • connected to a centralized node (terminal)

    • connection to tier-2

  • inter-WBAN

    • smartphone, PC

    • ad hoc

    • Wi-Fi, Bluetooth, …

    • connection to tier-3

  • beyond-WBAN

    • terminal center (cloud, server)

    • storage and analysis

    • cellular, WLAN

Contribution of the paper

  • technical requirements of WBANs

    • sensors

    • frequency bands

    • channel models

    • ultra-short range comm.

    • networking

    • safety and privacy protection

  • chip-based solution

WBAN applications for health monitoring#

applications

  • medical

    • wearable

    • implantable

  • non-medical

Medical examples

WBANs for various diseases:

  • protocols used: Wi-Fi, cellular, Bluetooth, ZigBee

  • use smartphones

Technical requirements for different applications

  • Note: 10-10 means \(10^{-10}\).

Exercise 96

Pick three applications from this table, e.g., ECG, imaging, and deep brain stimulation. Then discuss whether the individual performance metrics like data rate, number of nodes, etc, make sense.

Desired ranges of different performance metrics

Exercise 97

This table suggests that peak power consumption should be 30 mW and in sleep mode some μW. Discuss whether this make sense.

How long can such a device run on a LiR2032 battery assuming following duty cycles?

  • 1 %

  • 10 %

  • 50 %

WBAN sensor techniques#

Sensors convert physical parameters to electronic signals.

!define types WBAN sensor types (types) -- (physiological) (physiological) -- (invasive, implantable) (physiological) -- (non-invasive) (types) -- (biokinetic) (types) -- (ambient)

Exercise 98

Look at data rates of different biosensors. Why is the data rate of

  • a humidity sensor low?

  • an accelerometer high?

  • a blood pressure sensor low?

Wireless Transmission in WBANs#

Need for short-range, low-cost, low-power communication

  • generally 1 mW peak power desired

Frequency bands

  • medical device radio communications band (401-406 MHz)

    • used in implantable devices

  • ISM band 2360-2500 MHz

    • interference in 2400-2483 MHz occupied by Wi-Fi, Bluetooth, ZigBee.

  • human body communication (HBC) for using the body as a wire, e.g., body surface to body surface communication

  • According to the table which compares different bands critical applications like implants have their own band, however the bandwidth is limited

802.15.6 channel models

  • implant to implant

  • implant to body surface

  • body surface to body surface

  • body surface to external

Exercise 99

According to this table about different scenarios about different communication links on the body:

  • S1 and S2 use 402-405 MHz

  • S3 uses all available ranges

  • S4 uses ranges above 900 MHz

What could be the reason/s?

WBAN

  • does not suffer from multi-path fading

  • but path loss will change dramatically due to the movement of body parts

Path loss for WBAN

\(L_{path} = L_{path,0} + 10 n \log_{10}(\frac{d}{d_0}) + S\)

  • \(L_{path,0}\) path loss at a reference distance \(d_0\).

    • based on antenna’s size

  • \(d\) distance between the antennas

    • antenna’s position in the human body

  • \(n\) path loss exponent

    • height, weight, gender

  • \(S\) shadow

    • shadow effect occurs when the human body directly obstructs the line-of-sight path between devices. This creates a shadow zone behind the body

This is a static model and does not model a dynamic scenario like walking, running etc.

Physical layer (PHY) defines

  • the physical communication medium

  • how the signal is modulated on the medium

Medium access layer (MAC) defines

  • how concurrent access to the medium is coordinated

  • the structure of a data packet (called frame)

  • how participants are addressed

Multiple access to the physical medium by allocating specific frequency band, time interval, or signal coding scheme. *DMA stands for *-division multiple access:

https://upload.wikimedia.org/wikipedia/commons/8/85/FDMA%2C_TDMA%2C_CDMA.png

Fig. 69 Frequency-, time-, and code-division multiple access schemes. Fuqia stands for power, Koha time, Frekuenca frequency.
CC BY-SA 4.0. By KaltrinaMu. Source: Wikimedia Commons#

FDMA, TDMA, CDMA have all the same spectral efficiency, but have individual challenges based on the communication scenario.

Another multiple access scheme is carrier-sense multiple access (CSMA).

  • CSMA-CD (collision detection)

  • CSMA-CA (collision avoidance)

  • CSMA-CR (collision resolution)

https://upload.wikimedia.org/wikipedia/commons/1/1f/CSMA-CR.png

Fig. 70 Example of CSMA-CR (collision-resolution). B may continue communication, while A and C stop communication after they sense that another participant is speaking. In this example, the node that transmits most low-level signals may continue with transmission and thus has a higher priority than others.
CC BY-SA 4.0. By Roumanet. Source: Wikimedia Commons#

https://upload.wikimedia.org/wikipedia/commons/1/1d/Csma_ca.svg

Fig. 71 CSMA-CA with and without request to send (RTS) and clear to send (CTS). If RTS and CTS are used, a transmission must be allowed by a station by issuing a CTS after an RTS.
Attribution. By jjgarcia.tsc. Source: Wikimedia Commons#

Exercise 100

Wireless systems cannot transmit and receive at the same time, because otherwise they would receive their own signals during transmission (other signals would be dominated by their own signal).

Based on this, which CSMA schemes would be applicable to wireless systems?

802.15.6 Physical layer (PHY) schemes

  • narrowband (NB)

    • ~400 - 2500 MHz

  • ultra wide band (UWB)

    • 3500 - 10000 MHz

    • these GHz level frequencies are highly absorbed by water in our body (shadow effect)

  • human body communication (HBC)

    • 5 - 50 MHz

WBAN power consumption affected by PHY and MAC design, e.g.,

  • MAC layer can use from the PHY layer about channel conditions to make decisions about when and how often to transmit packets. In case of low SNR (e.g., shadowing), the transmission rate can be reduced to save energy.

Network layers build on top of PHY and MAC. They ensure communication between two local networks, e.g., two WBANs. Example WBAN requirement:

  • different quality of service (QoS) levels. QoS determines how a data packet should be delivered, e.g., low latency, high priority, etc. E.g., ECG data should be prioritized over image data, because if we lose some pixels of the image, the image could still be recognized. However losing some bytes of ECG may distort the waveform, potentially leading to misinterpretation by a doctor.

Security and privacy#

  • Private data must be protected

  • Usual security schemes cannot be used in WBANs, because WBANs are resource-constrained.

802.15.6 defines three security levels (2017-802.15.6 4.6 Security paradigm)

  1. no security

  2. authentication (identifying a communication partner)

    • includes data integrity due to MAC (data is not modified on the path)

  3. authentication & encryption (and hiding data from observers)

Exercise 101

In what scenarios would each security level be appropriate? Provide an example for each.

WBANs offer interesting authentication features

  • biometric data like individual heart-beat interval changes can be used as a key for encryption.

Using dedicated hardware (ASIC) instead of software achieves less power consumption and latency.

Energy efficiency#

Size of the devices is sometimes limited to a few cm

  • => battery size also limited

  • frequent changes not possible (implant) or uncomfortable

Solutions

  • energy harvesting (using solar, vibration, thermal, radio frequency as energy source)

  • using dedicated hardware (ASIC)

Exercise 102

Look at this table that lists performance comparisons of some WBAN ASICs (chips).

  1. Which ASIC has the best energy efficiency for data transmission?

  2. Which performance metric could be the reason for this ASIC’s high energy efficiency?

Contents