AHA Demonstrates Complete DVB-S2X Modem in SDR Platform

6 篇文章 2 订阅

AHA Demonstrates Complete DVB-S2X Modem in SDR Platform

By Amy Cearley | Tuesday, June 30, 2015

Comtech EF Data Corp.'s AHA Products Group announced today, the successful demonstration of a complete DVB-S2X Modem in the Ettus X310 USRP™ Software Defined Radio (SDR) platform. This modem is based on the AHA4709E2IQ IP core that integrates all pieces of the digital communication path from Ethernet to I/Q symbols (ADC/DAC interface). The modem, as implemented, can transmit and receive data at up to 72 million symbols per second and at data rates of up to 324 million bits per second. The implementation loss is extremely low with the performance of the AHA4709E2IQ IP core itself exceeding the performance of the DVB-S2X specification for some codes.

An advancement of the successful AHA4702 DVB-S2 IP cores, the AHA4709 DVB-S2X series of IP cores comes with added support for lower roll-off factors, higher and lower order modulations, finer-stepped MODCODs and Very Low Signal-to-Noise Ratio (VLSNR) codes. AHA IP cores can be customized for throughput and/or size constraints and implemented in either ASIC or FPGA .

"Our customers have been excited to see us expand our high-performance FEC expertise to include the rest of the digital communications path," stated Jeff Hannon, AHA Product Group's General Manager and VP of Engineering. "They are able to focus on other pieces of their system knowing that they have a highly-optimized, high-performance solution from AHA."

"Ettus Research is excited to partner with companies like AHA that are pushing the leading edge in SDR performance on our USRP platform. The X310 is an ideal platform to develop new technologies rapidly and we are excited to see USRPs used in one of the first DVB-S2X implementations," said Matt Ettus, President of Ettus Research.

AHA's Forward Error Correction (FEC) and communication solutions are used by commercial and government customers around the world. With the long schedules required for developing standards-based communication IP, AHA solutions greatly accelerate customer's time-to-market and lower development costs.

About AHA Products Group

For over 26 years AHA has been the global leader in Lossless Data Compression and Forward Error Correction and now supports crypto acceleration technologies. Based in Moscow, Idaho, the design center develops and markets ICs, PCBs, and IP Cores for commercial and government customers worldwide. AHA products are used in Network Transmission, WAN Optimization, Enterprise Storage, Satellite and Terrestrial Communications, and a wide variety of other systems where data throughput and integrity are vital. A recognized technology innovator, AHA has a reputation for exceptional product quality, reliability, outstanding customer support and is an AS9100 certified manufacturer. AHA Products Group is a part of Comtech EF Data Corp., a wholly owned subsidiary of Comtech Telecommunications Corp. (NASDAQ:CMTL). For more information, please visit www.aha.com.

About Ettus Research

Ettus Research™, a National Instruments (NI) company since 2010, is the world's leading supplier of software defined radio platforms, including the Universal Software Radio Peripheral (USRP™) family of products. By supporting a wide variety of development environments on an expansive portfolio of high performance RF hardware, the USRP platform is the SDR platform of choice for thousands of engineers, scientists and students worldwide for algorithm development, exploration, prototyping and developing next generation wireless technologies across a wide variety of applications. For more information, please visit www.ettus.com or contact us at PR@ettus.com.

深度学习是机器学习的一个子领域,它基于人工神经网络的研究,特别是利用多层次的神经网络来进行学习和模式识别。深度学习模型能够学习数据的高层次特征,这些特征对于图像和语音识别、自然语言处理、医学图像分析等应用至关重要。以下是深度学习的一些关键概念和组成部分: 1. **神经网络(Neural Networks)**:深度学习的基础是人工神经网络,它是由多个层组成的网络结构,包括输入层、隐藏层和输出层。每个层由多个神经元组成,神经元之间通过权重连接。 2. **前馈神经网络(Feedforward Neural Networks)**:这是最常见的神经网络类型,信息从输入层流向隐藏层,最终到达输出层。 3. **卷积神经网络(Convolutional Neural Networks, CNNs)**:这种网络特别适合处理具有网格结构的数据,如图像。它们使用卷积层来提取图像的特征。 4. **循环神经网络(Recurrent Neural Networks, RNNs)**:这种网络能够处理序列数据,如时间序列或自然语言,因为它们具有记忆功能,能够捕捉数据中的时间依赖性。 5. **长短期记忆网络(Long Short-Term Memory, LSTM)**:LSTM 是一种特殊的 RNN,它能够学习长期依赖关系,非常适合复杂的序列预测任务。 6. **生成对抗网络(Generative Adversarial Networks, GANs)**:由两个网络组成,一个生成器和一个判别器,它们相互竞争,生成器生成数据,判别器评估数据的真实性。 7. **深度学习框架**:如 TensorFlow、Keras、PyTorch 等,这些框架提供了构建、训练和部署深度学习模型的工具和库。 8. **激活函数(Activation Functions)**:如 ReLU、Sigmoid、Tanh 等,它们在神经网络中用于添加非线性,使得网络能够学习复杂的函数。 9. **损失函数(Loss Functions)**:用于评估模型的预测与真实值之间的差异,常见的损失函数包括均方误差(MSE)、交叉熵(Cross-Entropy)等。 10. **优化算法(Optimization Algorithms)**:如梯度下降(Gradient Descent)、随机梯度下降(SGD)、Adam 等,用于更新网络权重,以最小化损失函数。 11. **正则化(Regularization)**:技术如 Dropout、L1/L2 正则化等,用于防止模型过拟合。 12. **迁移学习(Transfer Learning)**:利用在一个任务上训练好的模型来提高另一个相关任务的性能。 深度学习在许多领域都取得了显著的成就,但它也面临着一些挑战,如对大量数据的依赖、模型的解释性差、计算资源消耗大等。研究人员正在不断探索新的方法来解决这些问题。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值