Wireless Body Area Networks: Efficient Coexistence by Interference Management and Avoidance

شروع رویداد
چهارشنبه ۲۲ فروردین ۹۷ ۱۲:۱۵
پایان رویداد
چهارشنبه ۲۲ فروردین ۹۷ ۱۳:۳۰
مکان رویدادتهران
اضافه به تقویم
Wireless Body Area Networks: Efficient Coexistence by Interference Management and Avoidance
برگزارکننده‌ی رویداد
تماس با برگزارکننده
۰۲۱-۸۸۶۲۸۹۴۰
گزارش
مهلت ثبت‌نام برای این رویداد به پایان رسیده است.
جزئیات
مهلت ثبت‌نام
قیمت (تومان)
تعداد

توضیحات بیشتر

 

مؤسسه بین المللی توسعه دانش فردای ایرانیان با همکاری دانشگاه تهران برگزار می کند:

 

سمینار آموزشی 

Wireless Body Area Networks: Efficient Coexistence by Interference Management and Avoidance

 

Recent advancements in wireless communications, microelectromechanical systems (MEMS) and integrated circuits have enabled low-power, intelligent, miniaturized sensors strategically placed in, on, or around, the human body. Networks of these sensors are called wireless body area networks (WBANs). One key use for WBANs is health monitoring, so they transport mission-critical and delay-sensitive data. Thus, a reliable WBAN needs to be highly immune to interference. Based on the IEEE 802.15.6 standard devised for WBANs, nodes in a single WBAN can avoid interference by using techniques such as time division multiple access (TDMA). However, these networks social nature and high mobility imply the need for multiple WBANs coexistence; so with their mobility, it is generally not feasible to allocate a global coordinator to enable coexistence. Moreover, with an increase in the number of WBANs that coexist within close proximity, communications can be severely degraded.

Thus, this presentation looks into the new intelligent radio spectrum allocation strategies not only to avoid interference, but also to make best-use of limited available spectrum for reliable communication in large-scale deployments. First, we introduce an adaptive and partially orthogonal scheme called Smart Spectrum Allocation (SCA), which allows for synchronous and parallel transmissions from nodes in coexisting networks. This allows the practical coexistence of these networks giving far less communications delay, higher throughput and less interference.

A node's traffic priority, packet length, received signal strength along with density of sensors within a WBAN are further incorporated in the SCA protocol to comply with real-world variations in data type and length. We later deploy an energy harvesting module for sensor radios in these networks and use the radio interference as an energy source. To better optimize resource allocation, a hybrid scheme incorporating graph-coloring with pair-wise clustering of coexisting WBANs is proposed.

Next, in order to incorporate variations in channel assignment, based on body-dynamics mobility within each individual WBAN and amongst WBANs, a prediction algorithm is proposed to feed in these dynamics and accordingly update resource allocation. This proposal maximises the resource usage and transmission rate as well as offering a convenient trade-off between spectral reuse, transmission range and outage probability. Simulations and detailed analysis verify that the proposed approach is robust to changing channel conditions, increase in sensor node-density within each WBAN, and an increase in the number of coexisting WBANs.

Finally, we propose a self-organization scheme for the coexistence of multiple WBANs, which is biologically inspired from the theory of pulse-coupled oscillators. This approach achieves distributed management of the available spectrum amongst coexisting WBANs, and avoids the need for global management. It allows coexisting WBANs to use delayed information from previous transmissions to adjust a collision-free TDMA schedule, avoiding interference across WBANs, for future transmissions. Simulation results show that our protocol achieves rapid convergence to an updated spectrum allocation for higher packet delivery ratio, despite very-limited shared information from coexisting networks, implying that overhead is significantly reduced on WBANs with constrained resources.


با حضور 

Samaneh Movassaghi, PhD

Research Fellow
Google

 

مدت و زمان بندی: چهارشنبه 22 فروردین 1397 - ساعت 12:15 تا 13:30 

 

آدرس محل برگزاری:

تهران،اتاق 801ساختمان جدید دانشکده مهندسی برق و کامپیوتر، پردیس شماره 2 دانشکده¬ فنی دانشگاه تهران

آدرس:تهران اتاق 801ساختمان جدید دانشکده مهندسی برق و کامپیوتر پردیس شماره 2 دانشکده¬ فنی دانشگاه تهران