In conversational AI assistants, SLU models are part of a complex pipeline composed of several modules working in harmony. Hence, an update to the SLU model needs to ensure improvements not only in the model specific metrics but also in the overall conversational assistant performance. Specifically, the impact on user interaction quality metrics must be factored in, while integrating interactions with distal modules upstream and downstream of the SLU component. We develop a ML model that makes it possible to gauge the interaction quality metrics due to SLU model changes before a production launch. The proposed model is a multi-modal transformer with a gated mechanism that conditions on text embeddings, output of a BERT model pre-trained on conversational data, and the hypotheses of the SLU classifiers with the corresponding confidence scores. We show that the proposed model predicts defect with more than 76% correlation with live interaction quality defects, compared to 46% baseline.

A major challenge encountered in the offline evaluation of machine learning models before being released to production is the discrepancy between the distributions of the offline test data and of the online data, due to, e.g., biased sampling scheme, data aging issues and occurrence(s) of regime shift. Consequently, the offline evaluation metrics often do not reflect the actual performance of the model online. In this paper, we propose online adaptive metrics, a computationally efficient method which re-weights the offline metrics based on calculating the joint distributions of the model hypotheses over the offline test data VS. the online data. It provides offline metrics which estimate the production performance of the model by taking into account the test data biases. The proposed method is demonstrated by real life examples on text classification and a commercial natural language understanding system. We show that the online adaptive metrics can provide accurate predictions of online recall and precision even with a small test dataset.

In the near future, it is envisioned that cellular networks will have to cope with extensive internet of things (IoT) devices. Therefore, a required feature of cellular IoT will be the capability to serve simultaneously a large number of devices with heterogeneous demands and qualities of channel state information at the transmitter (CSIT). In this paper, we focus on an overloaded multiple-input single-output (MISO) broadcast channel (BC) with two groups of CSIT qualities, namely, one group of users (representative of high-end devices) for which the transmitter has partial knowledge of the CSI, the other group of users (representative of IoT devices) for which the transmitter only has knowledge of the statistical CSI (i.e., the distribution information of the user channels). We introduce rate-splitting multiple access (RSMA), a new multiple access based on multi-antenna rate-splitting (RS) technique for cellular IoT. Two strategies are proposed, namely, time partitioning-RSMA (TP-RSMA) and power partitioning-RSMA (PP-RSMA). The former independently serves the two groups of users over orthogonal time slots while the latter jointly serves the two groups of users within the same time slot in a non-orthogonal manner. We first show at high signal-to-noise ratio (SNR) that PP-RSMA achieves the optimal degrees-of-freedom (DoF) in an overloaded MISO BC with heterogeneous CSIT qualities. We then show at finite SNR that PP-RSMA achieves explicit sum rate gain over TP-RSMA and all baseline schemes by marrying the benefits of PP and RSMA. Furthermore, PP-RSMA is robust to CSIT inaccuracy and flexible to cope with quality of service (QoS) rate constraints of all users. The DoF and rate analysis helps us in drawing the conclusion that PP-RSMA is a powerful framework for cellular IoT with a large number of devices.

We consider the problem of cache-aided interference management in a network consisting of KTsingle-antenna transmitters and KRsingle-antenna receivers, where each node is equipped with a cache memory. Transmitters communicate with receivers over two heterogenous parallel subchannels: the P-subchannel for which transmitters have perfect instantaneous knowledge of the channel state, and the N-subchannel for which the transmitters have no knowledge of the instantaneous channel state. Under the assumptions of uncoded placement and separable one-shot linear delivery over the two subchannels, we characterize the optimal degrees-of-freedom (DoF) to within a constant multiplicative factor of 2. We extend the result to a decentralized setting in which no coordination is required for content placement at the receivers. In this case, we characterize the optimal one-shot linear DoF to within a factor of 3.

The end-to-end learning of Simultaneous Wireless Information and Power Transfer (SWIPT) over a noisy channel is studied. Adopting a nonlinear model for the Energy Harvester (EH) at the receiver, a joint optimization of the transmitter and the receiver is implemented using Neural Network (NN)-based autoencoders. Modulation constellations for different levels of "power" and "information rate" demands at the receiver are obtained. The numerically optimized signal constellations are inline with the previous theoretical results. In particular, it is observed that as the receiver power demand increases, all but one of the modulation symbols are concentrated around the origin and the other symbol is shot away from the origin.

We consider the cache-aided MISO broadcast channel (BC) in which a multi-antenna transmitter serves K single-antenna receivers, each equipped with a cache memory. The transmitter has access to partial knowledge of the channel state information. For a symmetric setting, in terms of channel strength levels, partial channel knowledge levels and cache sizes, we characterize the generalized degrees of freedom (GDoF) up to a constant multiplicative factor. The achievability scheme exploits the interplay between spatial multiplexing gains and coded-multicasting gain. On the other hand, a cut-set-based argument in conjunction with a GDoF outer bound for a parallel MISO BC under channel uncertainty is used for the converse. We further show that the characterized order-optimal GDoF is also attained in a decentralized setting, where no coordination is required for content placement in the caches.

In this paper, we look at a wireless network consisting of K fully-cooperating transmitters serving K receivers, each equipped with a cache memory. Each node is equipped with a single antenna and transmitters have access to partial channel state information. For a symmetric setting, we characterize the generalized degrees of freedom (GDoF) up to a constant multiplicative factor. We further show that the characterized order-optimal GDoF is also attained in a decentralized setting, with no coordination during the cache content placement phase.

This work investigates the interplay of coded caching and spatial multiplexing in an overloaded Multiple-Input-Single-Output (MISO) Broadcast Channel (BC), i.e. a system where the number of users is greater than the number of transmitting antennas. On one hand, coded caching uses the aggregate global cache memory of the users to create multicasting opportunities. On the other hand, multiple antennas at the transmitter leverage the available CSIT to transmit multiple streams simultaneously. In this paper, we introduce a novel scheme which combines both the gain derived from coded-caching and spatial multiplexing and outperforms existing schemes in terms of delivery time and CSIT requirement.