Artificial Super Intelligence

We describe the throughput performance of user scheduling system when multiuser diversity is applied to the network.   If the number of scheduling users is K in the user scheduling network, the required feedback bits is reduced from B to B – log2 K where B is the number of feedback bits and K is the number of active users in the network.   The channel model of the uplink multiuser MIMO system is given by   y_k = H_k * x_k + n_k   where H_k is the wireless fading  channel vector for user k, x_k is the scalar transmit signal, and n_k is the noise signal vector.

This blog talks also about an wireless story. I am now preparing one technical paper to submit to an international conference and a journal, subsequently.

Noise modeling n 1 , ..., n NK are independent complex Gaussian noise terms with unit variance Fading Modeling We consider a block fading channel with iid Rayleigh fading from block to block. The channel coefficients are iid complex Gaussian with unit variance.

In wireless communications, either more spectrum or higher spectral efficient communication technologies can increase the data rate, which enables to service various multimedia contents. Since frequency spectrum is a natural resource, it is inherently limited so that it costs very high. Therefore, it is very important to find highly spectral efficient communication technologies. However, new good technologies requires higher complexity in terms of development, installation and so on. Therefore, our research goal is either to optimize the current technology and to find a new innovative technology. Beamforming is a scheme to overcome the limitation of the spectrum resource. Beamforming is used for spatial division multiple access which increases the system throughput by spatially diving multiple user signals. Spatially diving let multiple users communicate simultaneously in a same frequency band. When beamforming is applied for the downlink transmission, we call it as downlink beamform...

Tree diagram, or mind map, has been used for many areas including research, business, study. One of example programs to build tree diagrams is Freemind which is easy to use. While I use freemind, I suddenly felt that Freemind is needed to be upgrade with a new feature. Remind that tree diagrams in Freemind resembles not full structure but only a half part of a full tree. Recent half tree form mindmaps are convenient to show various divided facts effectively. However, it is not very effective to analyze facts and to tell a story about associated actions. Now, I propose the following new usage of tree based mind-map:

Nowadays, I have attended 3GPP meetings. It has been 3rd times since I first go abroad to join 3GPP in Shannhai last October. It is because the 3GPP is the end of its period that I become to think that the standardization is not quiet related to 'Advanced Research Area'. There are so many topics even in only RAN1 area. However, when it comes to the research topics in SAIT, only the very small area in RAN1 is related to what we have been doing in SAIT. Most topics discussed here are signaling issues related to controling format, how to report CQI/PMI, and rank something like that. These are quiet 'Standardization Issue', not the research issues what we have beending doing in SIAT. Before I leave Korea, my senior boss, VP asked me to remind the purpose of the business trip. However, whenever I have presentation for the business trip in front of him, he asked me more than I could and more than someone could get from 3gpp RAN1 standard meeting. I'm nowadays quiet confus...

What is network MIMO?

Up to now, wireless networks has been developed to make users compete each other. For instant, the user scheduler in the cellular network selects the user whose wireless channel is the best of all. Therefore, each user report his status as accurate as possible. He doesn't care other user when he report his channel quality. Even though his channel is good, if he continuously report his good channel quality, the other use whose channel is worse will receives much lower speed data. Consider the case the high quality user becomes in bad condition, his receiving performance will be also bad. Therefore, we need some cooperation to avoid this phenomena. One emerging way to avoid this inequality is to use cooperative diversity.