Handheld Consumer Devices Research Articles

Mobile 4K / 2K / HD video streaming supported by real-time FEC raptorQ codes

With the growth of packet-switch networks in the mobile consumer electronics market, fountain codes are coming to play an increasingly important role as they allow high packet loss rates while still supporting high QoS. This is particularly critical for video streaming in multimedia delivery on moving handheld mobile consumer devices. A powerful forward error-correction mechanism is the fountain coding scheme implemented by RaptorQ codes, defined over GF(256), whose decoding algorithms are very complex from a computational perspective, but highly suitable for parallelization. With multimedia delivery for handheld consumer electronics in mind, this paper proposes several decoding schemes for embedded mobile SoCs, exploiting both the parallel potential of the GPU and CPU, sharing a common memory space. The OpenCL and OpenMP frameworks are used to exploit both accelerators for the roles of matrix multiplication and inversion that comprise the bulk of computation in RaptorQ decoding. Reported results indicate low-power requirements within the range of typical handheld consumer devices (1.5~2.4W) for data processing rates of dozens of Mbit/s that are fully within the real-time requirements of HD, 2K and 4K coded video streaming resolutions and beyond.

Addressing Challenges in Automotive Connectivity: Mobile Devices, Technologies, and the Connected Car

With the dramatic mismatch between handheld consumer devices and automobiles, both in terms of product lifespan and the speed at which new features (or versions) are released, vehicle OEMs are faced with a perplexing dilemma. If the connected car is to succeed there has to be a secure and accessible method to update the software in a vehicle's infotainment system - as well as a real or perceived way to graft in new software content. The challenge has become even more evident as the industry transitions from simple analog audio systems which have traditionally served up broadcast content to a new world in which configurable and interactive Internetbased content rules the day.

Temporary Bonding Cost Of Ownership: The link between low total thickness variation and chip yield

The ability to process thin wafers with thicknesses of 20-50um on front- and backside is a key technology for 3D stacked ICs (3Ds-IC). The most obvious reason for thin wafers is the reduced form factor, which is especially important for handheld consumer devices. However, probably even more important is that thinner wafers enable significant cost reduction for TSVs. Consensus has developed on the use of Temporary Bonding / Debonding Technology as the solution of choice for reliably handling thin wafers through backside processing steps. Temporary bonding and debonding comprises several processes for which yield is essential, as costly fully functional device wafers are being processed. The presented temporary bonding process consists of a bi-layer system, a release layer, Dow Corning WL-3001 Bonding Release and an adhesive layer, Dow Corning WL-4030 or WL-4050 Bonding Adhesive, processed on EVG's 850XT universal temporary bonding and debonding platform. Furthermore, this bi-layer spin coated material allows a room temperature bonding-debonding process increase process throughput which translates to low cost of ownership for high volume manufacturing. As such, this bi-layer approach features high chemical stability exposed to phosphoric acid, nitric acid, organic solvents and other chemicals familiar to TSV fabrication. Besides chemical stability this adhesive system provides also a high thermal stability when exposed to temperatures up to 300 °C. The temporary bonding process yield has a major impact on the overall Cost of Ownership (CoO). On the other hand, throughput of the individual process steps like spin coating, bonding, cure, debonding and cleaning processes is the second determining factor for improved CoO. In this presentation, we will present a study of the total thickness variation (TTV) and the evolution of TTV at different stages of the process. High resolution in-line metrology is an enabling tool to trace the bond integrity and yield throughout backside processing. As TTV is a major determining factor of the overall process yield, understanding its impact over the bonded wafer pair carries major importance. Especially, non-continuity of the edge region, showing an inherent edge bead after coating, and edge die yield will be focus of our contribution. Finally, our experimental results will be transferred into a cost of ownership model, discussing the pros and cons for high volume production.

Exploiting digital TV users' preferences in a tourism recommender system based on semantic reasoning

Tourism recommender systems match the user preferences against the huge diversity of tourist resources, helping to decide where to go and what to do. Current approaches require the users to initialize manually their profiles by expressing their interests accurately, which is a very tedious process. We propose a system that automatically infers the users' preferences from their TV viewing histories, i.e., the tourism resources the users might appreciate are selected by considering the TV contents they enjoyed in the past. To this aim, we have developed a context-aware semantics-based recommendation strategy that considers both the users' preferences and the interests of like-minded individuals. The resulting recommendations shape a tailor-made on-move travel plan the users can access via (domestic and) handheld consumer devices.

Incentivized provision of metadata, semantic reasoning and time-driven filtering: Making a puzzle of personalized e-commerce

e-Commerce recommender systems select potentially interesting products for users by looking at their purchase histories and preferences. In order to compare the available products against those included in the user’s profile, semantics-based recommendation strategies consider metadata annotations that describe their main attributes. Besides, to ensure successful suggestions of products, these strategies adapt the recommendations as the user’s preferences evolve over time. Traditional approaches face two limitations related to the aforementioned features. First, product providers are not typically willing to take on the tedious task of annotating accurately a huge diversity of commercial items, thus leading to a substantial impoverishment of the personalization quality. Second, the adaptation process of the recommendations misses the time elapsed since the user has bought an item, which is an essential parameter that affects differently to each purchased product. This results in some pointless recommendations, e.g. including regularly items that the users are only willing to buy sporadically. In order to fight both limitations, we propose a personalized e-commerce system with two main features. On the one hand, we incentivize the users to provide high-quality metadata for commercial products; on the other, we explore a strategy that offers time-aware recommendations by combining semantic reasoning about these annotations with item-specific time functions. The synergetic effects derived from this combination lead to suggestions adapted to the particular needs of the users at any time. This approach has been experimentally validated with a set of users who accessed our personalized e-commerce system through a range of fixed and handheld consumer devices.

Robust DTW-based recognition algorithm for hand-held consumer devices

This paper presents a new robust dynamic programming (DP)-based recognition algorithm that is quite suitable for menu-driven recognition applications with small vocabulary size (typically less than 50). When compared to the conventional dynamic time warping (DTW)-based recognizer, the proposed algorithm shows significantly improved recognition accuracy in speaker-independent cases. In addition, when compared to the conventional hidden Markov model (HMM)-based recognizer, the proposed algorithm requires smaller computational amount and parameter file size, while maintaining almost the same recognition rate for command recognition applications with small vocabulary size in hand-held consumer devices.