Page 332 - From GMS to LTE
P. 332

318  From GSM to LTE-Advanced Pro and 5G

             For the calculation it was assumed that the device has a battery with a capacity of 5
            Watt hours. This is about a third of the battery capacity that is built into a smartphone
            today. The chapter further contains an assumption about how much power the device
            draws in different states. In the ‘idle’ state, the state that a device is in most often, power
            consumption is assumed to be 0.015 mW.
             If the device was in idle state all the time, the battery could power the device for 38
            years. This does not include battery self‐discharge, however. According to the Varta
            handbook of primary lithium cells [38], self‐discharge of a non‐rechargeable lithium
            battery is less than 1% per year and thus significantly less than the self‐discharge rate of
            rechargeable batteries.
             Obviously, a device is not always in idle state and when transmitting the device is
            assumed to use 500 mW of power. With this power consumption, the battery would
            only last 10 hours. As these two values are significantly different, the 3GPP study looked
            at different transmission patterns. If 200 bytes are sent once every 2 hours, the device
            would run on that 5 Wh battery for 1.7 years. If the device only transmits 50 bytes once
            a day the battery would last for 18.1 years.


            4.19.11  NB‐IoT – High Latency Communication
            One of the main requirements of many IoT scenarios is very deep in‐house coverage
            with a very low signal level. NB‐IoT addresses this issue by repeating data transfers
            many times to give the receiver the opportunity to combine the signal energy received
            during each transmission. The downside of this approach is that transmitting even a
            small IP packet takes a significant amount of time. An interesting calculation that can
            be found in an Ericsson paper on the topic shows that transferring a small UDP packet
            can require up to 7 seconds under very low signal conditions where the system repeats
            each individual transmission for system access, bandwidth assignment, user data trans-
            fer and acknowledgement several dozen times [39].
             Another main requirement of many IoT scenarios is the exchange of extremely low
            device power consumption for the sending and receiving of very little data and very
            long intervals in which no data is exchanged at all. If a device does not need to react
            instantly to incoming requests it does not make sense to periodically activate the radio
            module to check Paging messages. If, for example, it is enough to check once every half
            hour for incoming IP packets, the radio module can be completely switched off for most
            of this time, which saves a significant amount of energy. The downside is, of course, that
            in the worst case it takes 30 minutes for a device to respond to an incoming IP packet.
            To address such scenarios the 3GPP specifications were extended by a number of fea-
            tures for ‘High Latency Communication’, as described below.

            Extended Idle Mode Discontinuous Reception (eDRX)
            When a mobile device is in idle state it has to listen on the LTE paging channel for
            incoming Paging messages. These are sent when no active radio link is established and
            IP packets arrive from the Internet for the device. The device then answers the paging,
            a radio bearer is established and the IP packets are delivered. A typical paging interval
            in LTE networks today is 1.28 seconds, i.e. the radio chip of a device has to wake up once
            every 1.28 seconds and check the paging channel. While for smartphones the amount
            of power required to check for incoming Paging messages once a second is negligible
   327   328   329   330   331   332   333   334   335   336   337