Abstract: A method and system are provided for preventing data loss in a VoIP system. In particular, during a VoIP call, it is determined whether incoming ringing on a POTS line causes an unacceptable level of signal loss or errors. If so, for subsequent VoIP calls, the CO handling calls to the POTS line is instructed to either answer each call with a busy signal or automatically forward calls to the POTS line to the VoIP line or other selected telephone. Calling returns to normal upon ending of the VoIP call. In this manner, incoming ringing on the POTS line does not result in call dropping or lengthy retraining processes.

Abstract: The present invention is directed to methods and systems for implementing a DMA scheduling mechanism and a DMA system for transmission from fragmented buffers. According to an aspect of the present invention, a processor controls several devices via a polled interface to interleave DMA data transfers on different Input/Output (I/O) ports in an efficient manner. According to another aspect of the present invention, a system for handling transmission of network packets which are assembled from multiple memory buffers with different octet alignments is provided. The hardware/software combination allows efficient joining of packet fragments with differing octet alignments when the underlying memory system is word based, and further allows insertion of other data fields generated by a processor.

Abstract: A circular array structure is maintained having multiple time slots, where each time slot corresponds to a timeslice during which CPU resources are allocated to a particular thread. The time slots in the circular array include a queue of threads scheduled for execution during that time slot. A pointer index and an array of threads requesting immediate CPU resource allocation are maintained. A currently executing thread is suspended, and a next time slot during which the currently executing thread should resume execution is calculated. The suspended currently executing thread is appended to the queue of threads scheduled for execution at the calculated time slot. The pointer index is undated to point to the identified next sequential non-empty time slot. Any contents of the indexed time slot is appended to the array of threads requesting immediate CPU resource allocation. The thread at the top of the array is removed and activated.

Abstract: A method and apparatus for scheduling the transmission of cells onto an network, or other packet switching network, is disclosed. The central feature of the scheduling mechanism is a quality of service engine (QoS Engine) which accelerates the processing of packets in a packet switching networks, such as an ATM network, by assisting the accurate pacing of many ATM virtual circuits. The QoS Engine allows the concurrent support of a wide variety of port speeds, traffic classes using different priorities and traffic parameters. quality of service engine (QoS Engine) works in conjunction with a network processor (NP) to allow it to maintain software flexibility, and for it to achieve accurate pacing.

Abstract: A system and a method for constructing a signal integrity supervisor capable of both detecting and triggering an appropriate response when transmit path signals indicate a potential damaging transmitter operating mode. The system and method of the present invention takes advantage of the inherent property of a Delta-Sigma Modulator (DSM) which makes the probability of encountering a long string of consecutive ones or zeroes during nominal operation very small. The signal integrity supervisor ensures safe transmitter operation by monitoring the data and the clock inputs to a digital to analog converter. The system may comprise a data signal supervisor and a clock signal supervisor. The data supervisor may comprise a comparator and a counter and may be configured to power down a line driver upon detecting a data stream having a continuous voltage level.

Abstract: A system and a method for constructing a signal integrity supervisor capable of both detecting and triggering an appropriate response when transmit path signals indicate a potential damaging transmitter operating mode. The system and method of the present invention takes advantage of the inherent property of a Delta-Sigma Modulator (DSM) which makes the probability of encountering a long string of consecutive ones or zeroes during nominal operation very small. The signal integrity supervisor ensures safe transmitter operation by monitoring the data and the clock inputs to a digital to analog converter. The system may comprise a data signal supervisor and a clock signal supervisor. The data supervisor may comprise a comparator and a counter and may be configured to power down a line driver upon detecting a data stream having a continuous voltage level.

Abstract: The present invention is directed to methods and systems for optimizing UTOPIA CLAV polling arbitration. According to an aspect of the present invention, UTOPIA L2 CLAV status polling of each PHY address may be optimized by polling PHY addresses that have not yet indicated an active CLAV status or have just finished a cell transfer so that a CLAV response is required again. According to another aspect of the present invention, UTOPIA L2 CLAV status polling may be arbitrated so faster connection PHY addresses are polled proportionally more often than slower connection PHY addresses. According to yet another aspect of the present invention, if a connection speed no longer has any PHY addresses which require polling, the arbitration may be altered so only the connection speed with PHY addresses which require polling are actually polled. This ensures that the polling bandwidth is used as efficiently as possible.

Abstract: A method and system are provided for preventing data loss in a VoIP system. In particular, during a VoIP call, it is determined whether incoming ringing on a POTS line causes an unacceptable level of signal loss or errors. If so, for subsequent VoIP calls, the CO handling calls to the POTS line is instructed to either answer each call with a busy signal or automatically forward calls to the POTS line to the VoIP line or other selected telephone. Calling returns to normal upon ending of the VoIP call. In this manner, incoming ringing on the POTS line does not result in call dropping or lengthy retraining processes.

Abstract: The present invention overcomes various problems by defining two upstream masks (U1, U2) and two downstream masks (D1, D2) and using a mask selectable system for the long reach digital subscriber line (LDSL), in which a unique modem feature is activated during handshake to automatically check for physical layer status in terms of spectral compatibility and, thus, automatically optimize the boosted mode with the use of the mask selectable system choose the best combination of upstream/downstream masks in any physical layer noise scenario.

Abstract: The present invention is directed to methods and systems for allocating bandwidth (or other shared resource) among multiple masters. According to an aspect of the present invention, an arbiter assigns a bucket to each CPU (or other device) where each bucket holds the credits for that CPU. Each bucket has a predetermined fill rate and a drain rate. Depending on the priority given to a particular CPU, the corresponding bucket will drain (or fill) at a particular rate. For example, CPUs with a higher priority will drain at a slower rate. For each clock tick (or other period of time) that a CPU is stalled, a number of credits is accrued. The bucket with the highest number of credits has priority and will be given access to the shared resource (e.g., DRAM, SDRAM, SRAM, EPROM, etc.).

Abstract: The present invention is directed to implementing theoretical foundations of frequency domain methods aiming to locally decrease the impact of aggregate side lobes of a few discrete frequency tones, in the vicinity of other “victim” frequencies. The family of frequency methods referred to makes use of one or more “canceller tones.” The “canceller tones” loads are optimized to minimize the aggregate power in the neighborhood of the victim frequencies.

Abstract: Disclosed herein are various techniques for multilevel channel coding. For a given binary word comprising a plurality of bits, the least-significant bit may be encoded by a boosted coding technique, such as turbo channel coding or low density parity check (LDPC) channel coding. A subset of the remaining bits of the binary word may be encoded using trellis coded modulation (TCM) coding. The results of the boosted channel coding and the TCM coding may be mapped to one or more symbols. The one or more symbols then may be transmitted to a receiver using transmission techniques well known to those skilled in the art. The present invention finds particular benefit in digital subscriber line (DSL)-based systems in accord with the ITU G.922.1 Recommendation and similar standards.

Abstract: System and method for partitioning a DSLAM network. In this regard, one such network can be broadly summarized by a representative communication system comprising a digital subscriber line access multiplexer (DSLAM) that is communicatively coupled on its line-side to a high-speed digital link. The DSLAM is communicatively coupled through the high-speed digital link with the trunk side of a remote line access unit (RLAU). The RLAU is communicatively coupled on its line-side to a first digital subscriber line (DSL).

Abstract: A system and method for improved synchronous access of stored data are provided herein. A data requestor transmits a clock signal and a read request signal for reception by a data source, whereupon skewed versions of the clock signal and the read request signal are received due to the delays in the signal paths between the data requestor and the data source. Accordingly, the data requestor provides skewed clock and read request signals to its input sampling module to simulate the delays of the signal paths. Additionally, the data requestor provides process information associated with the requested data to a dual clock first in-first out (FIFO) buffer. When the input sampling module detects a read request using the skewed read request signal, the input sampling module can use this signal and the skewed clock signal to sample a data signal from the data source to obtain the requested data.

Abstract: Various embodiments for addressing the performance objectives of LDSL and examples of smart systems for LDSL are disclosed. An evaluation of the spectral compatibility of two LDSL modes based on two different downstream masks, identified herein as LDSL Wide and Narrow, is disclosed. Spectral compatibility is evaluated in accordance with existing rules. Other embodiments may further comprise determining features of upstream transmission.

Abstract: Disclosed are exemplary techniques for compressing ATM cells by substituting reduced headers for the original headers. A header table may be maintained at the transmitting end whereby headers are stored in table entries based on a table index generated from a hash function applied to the headers. A copy of the header table may be maintained at the receiving end and may be updated by populating idle cells with header table contents. Upon receipt of a reduced ATM cell at the receiving end, the reduced header may be replaced with a full header obtained from the copy of the header table maintained at the receiving end. The present invention further provides techniques for registration and management of hash collisions.

Abstract: A method and system for reducing the power to average ratio (PAR) at the transmitter and after the up sampling and interpolation filter. In the time domain, Fractional Local Peak Detection and Mitigation (FLPDM) locally (i.e. in the neighborhood where the high peak occur) detects the high peaks at a sampling rate of T/K and combines an optimal binary and a PAR-lowering sequence at the scale of one extended symbol for multi-carrier modulation and at the scale M of symbols for single-carrier modulation. Since the technique acts locally, FLPDM does not require peaks search and processing at the scale of many time samples, does not require iterative Fast Fourier Transform (FFT) and Inverse Fast Fourier Transform (IFFT) processing and it assures tunable millions of instruction per second (MIPS) performance.

Abstract: Mechanisms for providing a subscriber-side interface with a passive optical network are described herein. An optical network termination (ONT) having an integrated broadband passive optical network processor is utilized to receive downstream data from an optical line termination (OLT) via a passive optical network and provide the contents of the downstream data to one or more subscriber devices via one or more data interfaces. Similarly, the ONT is adapted to receive and transmit upstream data from the one or more subscriber devices to the OLT via the passive optical network. The ONT preferably implements one or burst buffers for buffering upstream and/or downstream data. The ONT can be adapted to notify the OLT of the status of the burst buffer, thereby allowing the OLT to modify the bandwidth allocations.

Abstract: An improved ADSL modem is presented. The ADSL modem in cooperation with a remote ADSL modem will exchange manufacturer identification information during initial DSL system training to establish a DSL communication link. The improved ADSL modem may comprise a memory device containing system parameters and operational algorithms previously determined to maximize DSL link performance between ADSL modems manufactured by different vendors. The improved ADSL modem is well suited for central office applications as it provides a solution for optimizing system performance that is transparent to remotely located ADSL modems. An ADSL modem in accordance with the present invention may comprise a digital signal processor and a memory device configured to store appropriate system parameters and operational algorithms that provide an optimized DSL communication link once applied during system initialization. The present invention also provides a method for optimizing DSL system performance.

Abstract: In an asynchronous digital subscriber line (ADSL) system comprising a central office High Speed ADSL Terminating Unit (HSTU-C) in bi-directional discrete multitone (DMT) communication with a remote High Speed ADSL Terminating Unit (HSTU-R), a method for improving handshake detection is provided by the present invention. The method comprises transmitting handshake signaling from the HSTU-C to the HSTU-R via a first subset of carrier sets at a first symbol rate and transmitting handshake signaling from the HSTU-C to the HSTU-R via a second subset of carrier sets at a second symbol rate, the second symbol rate being less than the first symbol rate.