2016/10/28(五) 14:20 - 曾煜棋 院長(交大資訊學院) - Intelligent Edge Computing for mMTC
Topic： Intelligent Edge Computing for mMTC
Date&Time：2016/10/28 (五) 14:20
Speaker：曾煜棋 院長 交大資訊學院
Location：清華大學台達館 R216
Abstract：
IoT (Internet of Things) has been predicted as the next wave of ICT. On the other hand, software-defined networks (SDN) are receiving more attention. These two things do not seem to much related at the first glance. However, we observe that the success of IoT and SDX will not only rely on the power of devices and clouds, but also on the intelligence of edge networks.
On looking at the IoT networking issues, we identify the following features: (i) massive connectivity, (ii) devices generating tiny but periodical packets for the most of the time, (iii) rare event-driven traffics with high priority at emergencies. Traditional random access mechanisms do not fit well for mMTC. This demands for new design of MAC and signaling protocols for IoT networks. We will discuss how we model this problem and propose potential approaches assuming the existence of intelligent edge computing.
Date&Time：2016/10/28 (五) 14:20
Speaker：曾煜棋 院長 交大資訊學院
Location：清華大學台達館 R216
Abstract：
IoT (Internet of Things) has been predicted as the next wave of ICT. On the other hand, software-defined networks (SDN) are receiving more attention. These two things do not seem to much related at the first glance. However, we observe that the success of IoT and SDX will not only rely on the power of devices and clouds, but also on the intelligence of edge networks.
On looking at the IoT networking issues, we identify the following features: (i) massive connectivity, (ii) devices generating tiny but periodical packets for the most of the time, (iii) rare event-driven traffics with high priority at emergencies. Traditional random access mechanisms do not fit well for mMTC. This demands for new design of MAC and signaling protocols for IoT networks. We will discuss how we model this problem and propose potential approaches assuming the existence of intelligent edge computing.