University of Southern California
The next generation of emergency responder systems will be based on Long-Term Evolution (LTE), as the use of commercial oﬀ-the-shelf components will lead to a dramatic decrease in cost. However, since Public Safety Organizations (PSOs) often need to operate in areas where there is no cellular infrastructure (either by design, or because the infrastructure is nonoperative after a natural disaster), it is essential that Direct Mode Operation (DMO), also known as Device-to-Device (D2D) communication is fully operational and reliable. At the current time, the D2D mode of LTE is still under development, and no proper testing has been done yet. A key requirement is testing in realistic channels; it is not suﬃcient to analyze performance with the 3GPP (or similar) channel models, as those are intended to compare diﬀerent systems, not to evaluate absolute performance and reliability.
The key goal of this project is thus to perform extensive measurement campaigns for D2D channels, in particular concentrating on channels that are most important for PSOs, namely (i) channels for vehicle-to-vehicle (V2V) communications, including convoys, and (ii) indoor-to-outdoor (I2O) channels, where one mobile device is outdoors at street level, and the other indoors, possibly at a higher ﬂoor. For those scenarios, existing measurements are missing critical components such as (i) suﬃcient number of measurements to provide statistical viability, (ii) directional channel characteristics (which are needed to evaluate multi-antenna terminals), and (iii) evolution of channel characteristics when the device moves on a trajectory.
Principle Investigator: Andreas F. Molisch
University of Southern California
To perform such measurements, and evaluate all the parameters of interest, we will build a new dynamic, directional channel sounder, operating in the frequency range mandated for LTE D2D operation, and possibly additional frequencies suitable for public-safety applications. The sounder will build on a successful prototype we built for real-time measurements in V2V communications, but (i) will provide better angular resolution, (ii) be capable of evaluating elevation angles and polarization (which is especially important for I2O measurements), and can handle much larger movement speeds (critical for V2V measurements). For the evaluation of the channel parameters (directions, delays, and amplitudes of multipath components), we will use an iterative maximum-likelihood estimator (RiMax), which exists in our lab, to provide super resolution, i.e., resolution and accuracy that is much better than the basic Fourier resolution, both in delay and angle. We will furthermore develop/modify an Extended Kalman Filter (EKF) for the tracking of multipath components as the devices are moving on trajectories. Finally, based on our extensive experience in geometry-based stochastic channel models (GSCM), we will derive a model that eﬀectively describes all the important eﬀects, including nonstationarities, in the system. We will make our channel models available to other researchers in the NIST program for their investigation.
These channel measurements and models will be combined with simulations of the full LTE D2D stack to provide realistic performance ﬁgures. This approach is advantageous for two reasons: (i) the LTE D2D standard is still under development; changes in the standard can be incorporated into the software simulation much more easily than in hardware (even SDR) implementations; (ii) at the same time the accuracy of such full-stack simulation is very high, as modern simulators include even simulation of hardware nonidealities (iii) the combination of simulations with recordings of measured channels allows to test diﬀerent system modiﬁcations in a fair and reproducible manner, while at the same time avoiding the serious drawbacks (unrealistic behavior) of 3GPP or similar channel models. We will ﬁnally explore proposed system modiﬁcations (concentrating on those that do not require modiﬁcations of the standard) that can enhance the reliability to meet PSO speciﬁcations.
The PI and his team have ample experience, as well as worldwide-unique equipment and evaluation software, for the proposed work, in particular in the area of channel measurement and modeling. As a matter of fact, they have over the past 20 years led many “ﬁrst of its kind” measurement campaigns, and provided models that are now in worldwide use in both industry and academia.