5G on the way

5G has been a rather hot research topic for a while and this trend stays unchanged or even gains more momentum in Globecom 2015 in San Diego. Indeed, there were 4 tutorials dedicated to 5G and a couple more related topics like green network, fog network, full duplex, quantum communications, etc. There’are also several workshops for 5G and many more technical 5G symposiums. Meanwhile, there’re also growing focuses on IoT, indoor networks, caching, etc.

Commercialization of technologies is getting faster and faster than ever, even results from fundamental research. It is not clear who is leading, academia or industry. Those who discover something first and commercialize fast will earn huge returns quickly. For example, in the presentation, “indoor mm-Wave Channel Measurements: Comparative Study of 2.9 GHz and 29 GHz”, Qualcomm reported similar measured channel properties of 2.9 GHz and 29 GHz in indoor environment and is still trying to figure out why! But measurement is measurement! Theory always need to be corrected to follow facts. If the reported results are true and may even be universal for a broad band of spectrum, a lot will have to be changed.

Peiliang and I presented our recent research results on “‘Area Spectral and Energy Efficiency Analysis of Cellular Networks with Cell DTX'”. The research on the SE and EE of the overall network has always been a challenging area because of its complexity in analysis and non-concavity from interference terms.
However we successfully obtained the network spectral efficiency and energy efficiency as functions of network traffic load.
It is shown that the network spectral efficiency increases monotonically in traffic load, while the optimal network energy efficiency depends on the ratio of the sleep-mode power consumption to the active-mode power consumption of base stations. If the ratio is larger than a certain threshold, the network energy efficiency increases monotonically with network traffic load and is maximized when the network is fully loaded. What’s more, the power ratio threshold depends solely on the wireless channel propagation, i.e. path loss exponent
α. For example, the ratio is 56% for α = 4, indicating that in this propagation environment, if the sleep power consumption is more than 56% of the active mode power consumption, the network should be always fully loaded to maximize network energy efficiency, as sleeping won’t save you much. Otherwise, there is an optimal load that the network should work on so that the energy efficiency is maximized.

This entry was posted in Uncategorized. Bookmark the permalink.

1 Response to 5G on the way

  1. Petri Mähänen says:

    Just commenting the indoor measurements on 2.9 GHz and 29 GHz. I try to be brief and not to become reviewer for the accepted paper. But basically you have to be careful to interpret the paper, and I guess that your main wondering is Figure 7 (path loss): The main similarity is on LOS-channel, but this is not surprising. You basically have short-range LOS, and in the environment where you will have PLE < 2 as you have a sort of wave-guide effects coming into the play (note that this is only for the received power, we are not talking about phase or delays here). Thus it is not too surprising that both frequencies have somewhat similar values; however, note that the sigma is quite a large in the plots (and more seriously we do not have good understanding on how much data overall has been collected), so we are somewhat limited statistics. You will notice that the situation is already somewhat different in NLOS, although perhaps there one could say surprisingly similar (but there could be specific location explanations). In the any case, if one is looking delay characteristics and such then we have clearly more differences.

    Rappaport et al. has been making rather large data set measurements, and gets for 28 GHz indoor LOS-channel PLE=1.7 for directional model(compared to this papers 1.4 for 29 GHz, and 1.6 for 2.9 GHz) and PLE=1.1 for omni-directional model (this is for V-V co-polarization; he finds PLE=4.1 for V-H cross-polarisation).

    But in general, as it is LOS indoor environment, this is not that much of surprising and we do not need to change our theories or insights.

    P.S. Note also that 29 GHz does not have oxygen absorption peak, so you would not expect to see 60 GHz problems in this band (and with this sort of distances).

Leave a Reply

Your email address will not be published. Required fields are marked *