Abstract: A new telecommunication service provides users with the ability to specify and receive spoken-language services in a preferred language that differs from a preferred written or text language. One or more user (e.g., a subscribing entity) preferences for spoken language as well as written language is specified and stored, e.g., in a subscriber database. An application server uses the configured spoken language preference(s) to select the language used for voice announcements, voice prompts, and other voice applications provided in one or more telecommunication services. For example, the server populates the existing SIP Accept-Language header and/or a new SIP P-Media-Language field with the spoken language preference(s) to allow the other servers, e.g., in a home and destination domain, to provide spoken/voice service as well as text-based service to the user according to the user's language preference(s).

Abstract: According to the teachings disclosed in this document, a receiver estimates impairment correlations for a received signal using received pilot values. This direct estimation of impairment correlations based on direct pilot observations, provides for a rich capture of all impairment sources bearing on the received signal. Advantageously, however, estimation noise is reduced by generating a structured covariance matrix from the measured covariance matrix, and carrying out structured filtering within the structured covariance matrix. Structured filtering involves averaging matrix elements that correspond to the same processing delay difference as used for measured impairment correlation estimation. In one or more embodiments, averaged net channel estimates are used to obtain a rank one matrix, which is used to convert a measured covariance matrix to a structured covariance matrix, for carrying out structured filtering.