由转载开始—通信—Mathworks

Engineering teams in the communications industry face the challenges of designing increasingly complex products in less time, while supporting evolving standards such as UMTS, LTE, WiMAX, TD-SCDMA, and VDSL.

These teams rely on MathWorks products to create innovative product designs faster. MATLAB^® and Simulink^® streamline design flows by helping them:

• Efficiently design communications algorithms and systems
• Combine digital, analog/mixed-signal, and control logic design domains with channel and traffic modeling, in a single system model
• Use the system model as an executable specification for simulation and prototyping
• Verify the implementation of the physical layer and MAC layer designs against the system model

Air Interface and Front-End Design

Air interface and front-end design teams use MATLAB and Simulink to quickly characterize how RF, digital signal processing, and other components interact, resulting in stronger engineering collaboration and faster system-level design iterations.

Cut Research Time in Half

Early exploration of the system behavior of analog/mixed-signal and RF is a time-consuming process when done with low-level tools. MathWorks tools enable early exploration of system behavior including mixed-signal and RF elements. Communications engineers model systems at a higher level of abstraction, incorporating digital signal processing and other domain-specific behavior when needed. This enables them to rapidly simulate the system with realistic environmental effects to improve the overall system design. The algorithms and other component models facilitate collaboration between the teams, and can be reused for implementation and verification.

For example, to improve an amplifier design while reducing component costs, engineers can model digital pre-distortion algorithms, compare them against analog implementations, verify the behavior using system simulations, and share them as reference models. These reference models help teams quickly verify the implemented algorithms in the final products.

Mixed-Signal Series: PLL Design

Bringing a Coherent System-Level Design Flow to AMS

Quickly Prove that Subsystems Work Together

Integrating analog and baseband subsystems with system-level models allows early identification of design flaws within subsystems and flaws at their interface boundaries. In the MATLAB and Simulink environment, communications engineers can connect subsystems modeled in different forms and different levels of fidelity (such as HDL models using Mentor Graphics^® ModelSim^®, C code for DSPs using Texas Instruments™ Code Composer Studio™, and analog circuit models using Cadence^® Virtuoso^®). Thus, they are able to ensure that subsystems will work together and that the designs match the original specifications.

Once the system model is created, it can be used as a test bench and reference model for the subsystem implementation. The test bench and input/output data can also be used with test equipment to verify the prototype or the final product.

BridgeWave Communications Delivers High-Capacity Wireless Links Months Ahead of Schedule

Early Verification of Digital-Analog Designs Using System-Level Cosimulation

Meet Deadlines and Quality Goals

Aggressive development schedules can compromise the quality of the air interface implementation. With a common design environment based on MATLAB and Simulink, RF and digital engineering teams can do faster iterations to improve designs and minimize errors, while keeping the development project on schedule.

Cambridge Consultants Develops WiMAX Test Bench for Aspex Semiconductor

Return on Investment in Simulink for Electronic System Design

 

Baseband Signal Processing

Engineering teams use MATLAB and Simulink to manage the complexity and scale of baseband signal processing projects.

Develop Algorithms in a Fraction of the Time

Today a vast majority of engineers use MATLAB to develop baseband signal processing algorithms in a fraction of the time it takes with C/C++ and other languages. They save time by working at a high level of abstraction and drawing from an extensive library of prebuilt signal processing algorithms such as FFTs, filters, and modulators. This makes it easy to develop baseband models based on OFDM or QAM and adapt to evolving standards such as UMTS, WiMAX, and LTE.

C-COR Cuts DSP Development Time by 30% for Cable TV Subsystems

Creating an Executable Specification for the WiMAX Standard

Ensure Signal Processing and Control Logic Subsystems Work Together

With MATLAB and Simulink, engineering teams can express both signal processing algorithms and control logic in a single system model. Engineers can describe the signal processing algorithms as MATLAB code, Simulink blocks, or custom C or HDL code, and include control logic expressed as MATLAB code or state machines.

The complete system is simulated to ensure that the different subsystems work together as specified. Using MathWorks parallel computing technology, engineers can also take advantage of multicore computers and compute clusters to accelerate long simulations for BER testing and parameter sweeps.

Engineers can reuse the system model as the test bench by introducing implementation-level subsystem models (in C, HDL, or analog circuits) in the simulation. The results can be directly compared with the system model simulations without additional scripts or test benches. The reference models can also be used to generate test vectors for verifying hardware prototypes. By connecting the hardware prototype and test and measurement equipment to the reference model, engineers can send test vectors to the prototype, capture the results, and compare them with the simulation results.

Realtek Semiconductor Cuts Development Time by 50% and Takes the Lead in New-Generation HDA Codecs

ETRI Develops Modem Synchronization Technology for 4G Mobile Telecommunications System

Speed Design Cycles While Increasing Quality

Designing at a high level of abstraction can dramatically improve the quality of baseband signal processing design. Using MATLAB and Simulink, engineers quickly model and simulate design ideas. This means they can perform more iterations to improve designs and uncover design flaws.

These engineers work at multiple levels of abstraction to test both concepts and implementation details. In contrast to design flows based on lower-level programming languages, MATLAB and Simulink enable engineers to combine the extensibility of a widely-used technical computing language and the ease and productivity of a higher-level system design tool, using different levels of design abstraction as needed. Additionally, engineering teams improve the verification of baseband signal processing systems by quickly and repeatedly verifying system designs – first against the original specification, and then against the final implementation – by utilizing links to external simulators, hardware prototypes, and test and measurement equipment.

Harman Becker Cuts Verification Time for OFDM Radio Receivers by 75%

Return on Investment in Simulink for Electronic System Design

Wireless and Wired Channel Modeling

To simulate and verify mobile phones, base stations, and optical network equipment, communications engineering teams need a comprehensive set of wireless and wired channel models. They use MATLAB and Simulink to develop these channel models for both desktop simulation and hardware prototyping, so they can analyze the behavior of the communications system under various impairment scenarios.

Verify Communications System Designs by Using Channel Models

MATLAB and Simulink give engineers a head start, providing prebuilt channel models that can be integrated into the system model for simulation under various impairment scenarios. Engineers can also customize channel models easily to meet their specific needs.

As a result, equipment providers are able to verify designs sooner and reduce the need for building and testing with hardware prototypes and network test equipment.

Elektrobit Testing Ltd. Saves a Year in Development Time for High-Resolution Radio Channel Measuring System

ETRI Develops Modem Synchronization Technology for 4G Mobile Telecommunications System

Running Channel Models on FPGAs

MIMO channels are complex to model and difficult to simulate. As a result, 4G or next-generation communications system and protocol simulations can take several days or even weeks to run. Engineering teams accelerate these simulations by using the power of FPGAs together with MathWorks tools, and third-party tools such as Xilinx^®System Generator and Altera^® DSP Builder.

From MIMO channel models developed in MATLAB and Simulink, these engineers automatically generate HDL code and implement the channel model directly on an FPGA board to emulate the MIMO channel effects for hardware-in-the-loop simulation.

Introduction to MIMO Systems

Rapid Design and Implementation Using Automatic HDL Code Generation

Traffic Modeling and Network Performance

Engineering teams use MATLAB and Simulink to model and simulate network traffic behavior and predict network performance – a critical need and a competitive advantage for designing and deploying the next generation of communications services and networks.

Design Networks and Services with Confidence

New communications networks must be able to transmit different services such as audio and video with a consistent level of quality. Network operators use MathWorks products to model and predict network performance measures such as congestion, resource contention, and process delays.

Video Streaming Over a Bandwidth-Limited Network 8:35

Modeling Discrete-Event Systems

Use Traffic Models to Predict Network Equipment Performance

Network equipment manufacturers use MathWorks products to create accurate and reusable traffic models of communications equipment and protocols to simulate network and channel utilization, as well as the impact of adding new services on existing networks.

Engineers use traffic models to determine latency, blocking, and resource starvation issues inherent in a communications network configuration. These models allow equipment manufacturers to demonstrate and prove the performance and suitability of their products to network planners and service providers. Network planners use the models created by equipment manufacturers to study and compare various network configuration options, and recommend the most efficient solutions to service providers.

SimEvents for Architecture Exploration Based on Performance Analysis 7:45

Ethernet Bus Modeling 4:19

Model Physical and Network Layers in One Environment

Engineers need to simulate and verify the network performance while the physical layer is still under development, or independent of the physical layer behavior. To help simulate network performance independently, MATLAB and Simulink support both time-driven and discrete-event modeling including queues, servers, and traffic generators. Engineers also use the tools to model the network performance and the physical layer together. To do this, they combine the network traffic models with physical layer components including state machines, signal processing, and mixed-signal subsystems, simulating both the discrete-event and dynamic system behaviors.

Discrete-Event Simulation with SimEvents 2:28

Advanced Modeling Techniques for Event-Based Systems