Ray joined Award Solutions in 2006, bringing his expertise in CDMA technologies and overall experience in real-time product development, delivery and support of wireless telecommunications systems. He has over 20 years of experience in the wireless telecom industry. At Award Solutions, he specializes in 4th generation telecommunications systems, with a focus on WiMAX technology, engineering and deployment.
Ray began his career working on real-time embedded software development in data communications for computer numerical controls. He then joined Nortel and worked on IBM-compatible data terminal systems development and other proprietary data communications systems for 8 years. He then moved into development for class 5 switching systems and followed that with more than 16 years in various development and support roles in wireless telecommunication systems.
As a senior designer at Nortel, Ray worked on the development of real-time call processing and OA&M software development for AMPS and TDMA. As a manager and senior manager for CDMA development, he managed teams that developed and delivered the first deployment of the CDMA IS-95B BSC product and also delivered products based on the CDMA2000 1XRTT standard. He has also worked on stress test tool development for testing and analysis of link capacity in wireless networks. He has also worked on OA&M development and deployment tools for Nortel’s Media Gateway product. Ray has received numerous awards within Nortel for his work on CDMA new product development and performance improvements.
Ray’s position at Award Solutions puts him at the forefront of emerging technologies with an emphasis on WiMAX technology and deployment. He has extensive knowledge of Mobile WiMAX standards and hands-on experience with engineering and deployment teams. As part of Ray’s expertise in WiMAX technology, he was awarded a Qualified Specialist certification from Cisco as an approved installation, commissioning and field engineer for their BWX WiMAX product. He is also an approved instructor for Cisco’s BWX Broadband Wireless Access System.
Ray received his Bachelor’s degree in Computer Science (with Distinction) from Wayne State University in Detroit, Michigan.
This is part 4 of my blog series about PDSCH transmission modes for the UE’s C-RNTI.
Here we talk about transmission mode (tm3)…the first of the four MIMO schemes used in LTE 3GPP release 8. Tm3 has a few names: Open-Loop Spatial Multiplexing, Open-Loop MIMO, Large Delay CDD, or just MIMO (for those not concerned about the details of how….. You know those people: “Just give me my <insert favorite social networking or sports site name here> updates NOW!” with nostrils flaring and their smart phone clutched in a death grip).
Note the transmission scheme in the partial table below (copied from 3GPP TS 36.213).
A UE configured for tm3 will either support the robust and reliable “transmit diversity” or a version of high throughput MIMO named “Large-delay cyclic delay diversity” or “Large-delay CDD”. It will support MIMO for 2, 3 or 4 transmission layers.
Transmission mode 3 has the best of both worlds: support for robust transmission and high throughout MIMO without all the complexities of the other 3 MIMO transmission schemes. So this tm will support fixed and mobile applications in addition to machine-to-machine apps in the same cell. This mode doesn’t help category 1 UEs much since they don’t support DL MIMO.
Think of tm3 as the entry point for MIMO. It is probably the simplest MIMO transmission scheme to support as it doesn’t require as much feedback from the UE (less complexity in the UE) as it does for the other MIMO schemes and the decision-making is reduced in the eNB. Less MIMO work to do for all!
Except…<insert dramatic pause>.....now the eNB needs to make real decisions on what transmission mode to use for each individual UE. But it is probably simpler than it sounds. Based on the CQI and RI values (no PM feedback required) received from the UE, the scheduler software in the eNB will dynamically choose whether to transmit using the transmit diversity scheme or the large-delay CDD scheme. So even if your device supports MIMO, it won’t get MIMO transmissions unless the estimated DL channel quality is above a certain threshold and Rank Indication is greater than 1.
The UE configured for tm3 will do similar feedback reporting as tm1 and tm2: CQI and RI, except the CQI values will be based on the reported RI...so a reported RI of 1 means report CQI based on transmit diversity transmission else report CQI based on a transmission using large-delay CDD transmission scheme with 2, 3 or 4 transmission layers.
Since no PMI feedback is reported, the precoding matrix value for transmit diversity (on 2 or 4 antennas) is a defined formula in the 3GPP standards. For large-delay CDD, they use a known delay value with a known precoding matrix value for 2-antenna transmissions and for 4 antennas they use a pre-defined subset of up to 4 precoding matrices depending on the number of transmission layers used (i.e. use a different precoding matrix on each transmission layer). While this means less decision-making by the eNB and less complexity in the UE (no PMI measurements), your PMI values aren’t based on the current channel conditions so MIMO performance may not be consistent in different channel conditions.
Since MIMO is supported by tm3, the UE now needs to start looking for another DCI in the DL control channel: DCI Format 2A. DCI format 2A records have the additional fields to support transmission of 2 codewords (only one for transmit diversity) and also inform the UE of how many layers for large-delay CDD so it knows which precoding matrix values will be used for the DL MIMO transmissions. Format 2A also lets the UE know whether 2 or 4 antennas are being used for transmit diversity.
Transmission mode 3 is your entry-level MIMO so you can enjoy the benefits of higher speed data (when channel conditions allow) without a lot of additional feedback and still get the benefits of robust transmit diversity (when channel conditions require it).
Thanks Ray for this article. I have a question regarding TM4. Consider an eNB having two antenna ports. With TM4, now the scheduler has 3 options
1. Tx diversity(1 codeword mapped to 2 layers)
2.Spatial multiplexing(2 codeword mapped to 2 layers)
3.(not sure)Spatial multiplexing(1 code word mapped to 1 layer)
Question>Are all the above three valid possibilities for TM4?If yes, then how should the scheduler decide between these three?