Multi-RAT Small Cells: An Efficient Solution for 3G/LTE Deployment

Multi-RAT Small Cells: An Efficient Solution for 3G/LTE Deployment

Huan-yu Su, Ph.D.Vice PresidentTechnical Marketing & B/DMindspeed Technologies, Inc.

1

Nasdaq: MSPD

议程 (Agenda)

Mindspeed公司简介 (Company Overview)

无线移动网络的现状与今后的挑战 (Current Situation of

Wireless Networks and Future Challenges)

Mindspeed推荐的解决方案：多模小基站 (Mindspeed

Recommendation: Multi-RAT Small Cells)

多模小基站的特性及组网方案探索(Multi-RAT Small Cell Characteristics and New Network Topology)

总结(Conclusions)

2Company Confidential

Nasdaq: MSPD

A History of Innovation in Southern California

33

Nasdaq: MSPD

The Rockwell Legacy

44

M-D-YYYY

Mindspeed: A History of Market Leadership

A Legacy of Innovation

Semiconductor Systems

5

Skyworks

Nasdaq: MSPD

Mindspeed Technologies, Inc. NASDAQ: MSPD

Leading communications IC company; public since 2003

History of innovation and technology leadership in new infrastructure markets

Over 550 employees worldwide; more than 360 engineers

6

The Industry Leader in Carrier VoIP Solutions

Nasdaq: MSPD 7

Mindspeed’s approach has been proven over 12 years across 1000s of

designs. We bring our proven expertise from multi-core wireline service

provider infrastructure to mobile networks.

SoCs are application-specific, and we have delivered SoCs for

demanding core telecom network applications … in VoIP, in fiber CPE …

and now in mobile.

Proven SoC Leadership

Nasdaq: MSPD

Systems-in-a-Chip (circa 1995)

8

Nasdaq: MSPD 9

SoC Platforms – The Rise of Multi-Core Computing

3rd generation Mindspeed VoIP SoC (circa 2005) – over 300M transistors in 90nm

6x VLIW / 64bit DSP

2x 32bit ARM CPU

EMAC

Multi Mb SRAM

UART

USB Host

IPSEC

Embedded

Cores:1 2 3

4 5 6

1

2

TDM

UTOPIA

Ethernet

MII/GMII/RGMII

PCI

DDR

USB

Flexible

Interfaces:

M-D-YYYY

Media Processing Technology LeadershipWith more than 100 issued US patents

Multi-core Architecture Leadership

100M ComcertoVoIP Ports Shipped

300

Mindspeed Achievement In VoIP

M-D-YYYY




300


1 DSP core4-channel VoIP

8 DSP cores + 8 Co-processors640-channel VoIP

M-D-YYYY




300


Same binary code across

more than 20 devices

Benefits to OEMs: TTM, Design reuse, Lower SW maintenance cost and less chance for bugsBenefits to Operators: Same quality, easy SW upgrade across different generations of equipment, unified enhancement services across entire network, etc.

More flexibility with enhanced path to maturity, yet a significantly reduced total cost of ownership

Nasdaq: MSPD 13

Block Diagram for VoIP and Data Routing Software Stack

MSPD Supplied Software CSP Supplied Software CSP Customer Software

PC

I D

river

Virtual Ethernet driver (control, data)

Host Kernel (Linux) including packet filtering, crypto API

T.38

FOIP

V.27,

V.29,

V.17

Networking and Routing Stacks

(IP,TCP,UDP, PPP, HTTP,ICMP,IPSec etc)

Shared Memory Interface driver

Open Source Router Code (e.g.Linksys)

RTP/RTCP or CPS

Enet

Driver

WAN ATM

Driver

(WAN

Utopia)

Dual Port

Seri

al D

river

USB D

river

Hard

ware

Cry

pto

Module

s

TDM Driver

Voice Packet classifier & switching/bridging

MPoA

G.168 Echo Cancel

G.711,729a/b/e

G.726,723a

MPoFR

AAL5 FRF.12

SPI D

river

Enet

Driver

LANCaller

ID G

en &

Det

DT

MF

Gen &

Det

TDMSignaling

POTS

Signaling

Packet Signaling(eg SIP, H.323)

Comcerto Device Driver

Comcerto Channel Module

PBXSwitching

Asterisk Open Source PBX

HDLC

Driver

(WAN

HSSI)

Eth, PPP Framing, IP, UDP Framing

Carrier Class SoC: Over 2 million lines of software code!

We are not only a SoC chip vendor, but also a system solution vendor

Nasdaq: MSPD

过去典型产品开发人员分布 (The Old Design Paradigm)

14

10 产品研发工程师 (Design

Engineers)

1 产品质量验证工程师(Testing Engineers)

2 产品测试工程师 (Verification

Engineers)

4 软件开发及维护团队 (SW

Developers)

Nasdaq: MSPD

现在产品开发的大概趋势 (The New Design Paradigm)

15

10 产品研发工程师

10 产品质量验证工程师

200 软件开发及维护团队

20 产品测试工程师

Nasdaq: MSPD

Where Do We Go Next?

16

Mindspeed is applying its SoC design expertise to challenge established wireless component leaders in the 4G/LTE base station market.

Nasdaq: MSPD

无线移动网络的现状 (Current Situation of Wireless Networks)

早期主流2G (GSM, CDMA 及TDMA)无线移动网无疑都是以宏基站(BTS) +TDM网络组网

- 得益于较低的工作频点(美国800MHz, 中国、欧洲及其它地区的900MHz)，基站较大的覆盖面积及良好的室内穿透能力都为早期低成本推动无线移动产业的发展起到不可低估的作用

高频工作频点的引入(美国1900MHz, 中国、欧洲及其它地区的1800MHz)，及其带来的一系列问题，第一次给业界敲响了警钟：

- 高频意味着基站覆盖面积的缩小• 提高发射功率夺回失去的覆盖？功耗及运营成本的提高

• 增加基站数量？新的选址和建设成本的挑战

- 高频意味着室内穿透能力的明显下降• 把射频天线移到室内室分天线的使用


Nasdaq: MSPD

无线移动网络的现状 (Current Situation of Wireless Networks)

Without exception, early 2G (GSM, CDMA or TDMA) deployment were all based on Macro BTS + TDM configurations

- Thanks to the low frequency bands (e.g., North America: 800MHz, China and Europe: 900MHz), large cell radius and good in-building penetration helped to grow the wireless industry with low deployment/maintenance cost

The needs for additional bandwidth pushed the industry to use higher frequency bands (e.g., North America: 1900MHz, China and Europe: 1800MHz):

- Higher frequency bands mean smaller cell coverage area• Increase the antenna emission power to increase coverage area？ Power

and cost increase

• Increase the number of BTS? Challenges of cell site acquisition and CAPEX

- High frequency bands also mean reduced in-building penetration• Move the antennas in-side the building Distributed Antenna Systems (DAS)


Nasdaq: MSPD

无线移动网络的现状-续 (Current Situation of Wireless Networks – Cont.)

3G工作频点的进一步提高(2.1GHz)无疑使已有的挑战更加炽热化

移动数据业务的推广及智能终端的普及加重了无线接入网络资源的压力

- 终端的吞吐能力在飞快地增长，从早期的单一语音业务(~10Kbps)至GPRS的128Kbps和EDGE的384Kbps，到目前W-CDMA的~10Mbps及今后LTE的~100Mbps

- 业界普遍的收费模式是包月，限累积吞吐量、但不限流量

- 今后的单一终端可以随时并且长时间地吞掉整个覆盖区域的容量，这会是一个非常难对付的问题覆盖区域小型化, 从而提高单位覆盖面积里的容量(或可以叫容量密度)，是不可避免的趋势


Nasdaq: MSPD


The introduction of 3G at 2.1GHz undoubtedly enhanced the challenges wireless industry is facing

Further development of wireless data services and the rapid adoption of smartphones increase the demand on wireless access capacity:

- UE capability is continuously increasing: from early voice only (~10Kbps), to GPRS (128Kbps) and EDGE (384Kbps), then to the current W-CDMA (~10Mbps) and future LTE (~100Mbps)

- Current pricing plans used by operators are all based on total monthly cumulative throughput, i.e. limit on volume but not throughput rate

- Therefore, a future single UE can perfectly consume the entire throughput capacity of a cell sector: a major challenge. Use of smaller cell size is an efficient and unavoidable approach


Nasdaq: MSPD

运营商已经被迫开始商用小基站- 日本软银大力推广使用W-CDMA家庭小基站，主要目的在于解决室内覆盖，以及降低对宏基站吞吐容量需求的压力

由于可运维和降低成本的需要，加上核心技术(如光纤)和产业链(如CPRI标准)的成熟，射频拉远和集中式基带资源池变为可能

- 但近年MIMO以及64-QAM的引进，又导致了基带设备(BBU)与拉远射频头(RRH) 之间传输的巨大压力

• 例如20MHz LTE 2x2能够承载的业务流量为150Mbps，但到CPRI接口就变成了大于2Gbps的I/Q流量，整整13倍之多

LTE的工作频点将在2.6GHz- 与3G的2.1GHz 相比较，基站覆盖能力将会有至少4dB以上的衰减

- 利用现有的2G/3G宏站站址和发射功率，必然导致大量覆盖空洞和盲点



Nasdaq: MSPD


Operators have already started to deploy Femtocells- Softbank and AT&T are deploying large volumes of W-CDMA Femtocells,

to improve in-door coverage as well as off-loading traffic from Macrocells

Due to cost and maintenance --- and benefiting from mature core technologies (e.g., CPRI) --- centralized BBU with RRH became possible

- However, with the introduction of MIMO and 64QAM, the traffic between BBU and RRU is dramatically increasing

• For example, while the throughput for a 20MHz 2x2 MIMO LTE is only 150Mb/s, the I/Q samples that need to be carried over CPRI are more than 2Gb/s

The frequency bands for LTE are likely to be at 2.6GHz in China- There is an additional 4dB coverage reduction compared with 2.1GHz 3G

- Using the current cell-sites and antenna emission power, there will be a large number of coverage blind spots


Nasdaq: MSPD

现有组网方式用于LTE有待解决的明显问题 (Major Problems For LTE Deployment Using Current Network Topology)

不断地增加宏基站站址或发射功率来解决高频带来的覆盖问题已经不再是可选的方式

- 合理解决覆盖补盲和传输的压力是业界必然面对的挑战

覆盖区域小型化，从而提高整网容量的趋势也不太可能再援用增加宏基站的方式实现

- 成本、功耗等等都是问题

室分天线尽管可以有效地解决室内覆盖的问题- 天线数量的增加改善了覆盖，但并没有提高吞吐容量

- 况且代价是非常高昂的• 因为得使用宏基站等级的设备

• 高频模拟信号的衰减大量浪费的功耗


Nasdaq: MSPD

现有组网方式用于LTE有待解决的明显问题 (Major Problems For LTE Deployment Using Current Network Topology)

Adding cell sites or increasing the antenna emission power are no longer viable solutions

- Efficiently providing coverage and reducing traffic demand between the BBU and RRU are challenges the industry has to overcome

Throughput density increases can only be achieved by cell-size reduction. This can not be effectively addressed using Macrocell site increases.

- Cost and power are all issues

Although DAS can effectively improve indoor coverage, it does not address the demands of throughput density

- The cost is very high, since Macrocell equipment is used

- Radio frequency analog signal attenuation rapidly overall power is very high


Nasdaq: MSPD

一箭多雕的解决方案：多模小基站 (Multi-RAT Small Cells: An Ideal Solution)

高频下的盲点—覆盖问题(Coverage blind spot due to high frequency bands)

覆盖区域小型化—容量密度提升(Throughput density improvement via miniaturization of cell sector)

今后CPRI的传输压力 (Major

transport issue between RRU and BBU)

室内覆盖及容量密度的提升 (In-door coverage and throughput density improvement)

多网协同覆盖及运维 (Multi-

RAT/Hetnet support and operation)


多模小基站(Multi-RAT Small Cell)

Nasdaq: MSPD

典型小基站的特性与优点 (Typical Multi-RAT Small Cell Characteristics)


体积小(small size): 选址灵活、布网便利 (easy to deploy)

功耗小(Low power): 可以实现以太网供电 (PoE)，进一步简便了布网及运维的难度和代价 (Support PoE, reducing deployment challenges and CAPEX/OPEX)

射频功率小 (Lowemission power):

覆盖半径不大，容量密度明显大于宏基站 (small coverage area,therefore higher throughput density)

用于室内 (In-door use): 不但解决覆盖问题，同时增加容量(Not only address coverage, but also increase throughput density)

低成本 (Low cost): 用单位吞吐容量计算，成本明显低于宏基站 (much lower cost/throughput density)

传输压力小 (Low backhaul cost):

回传是IP接口(而不是CPRI) (IP traffic instead of I/Q over CPRI)

软件平滑升级 (Easy SW upgrade):

如果使用合理的硬件、软件结构 (If appropriate HW/SW solutions are used)

支持多模 (Multi-RAT): 能更有效地解决或补充多网协同覆盖的问题 (More effectively supports multi-RAT or HetNet applications)

Nasdaq: MSPD

Transcede® SoC Dramatically Reduces System BOM

$$

… while significantly accelerating Time-to-Market

App I/O

App Hardware

Accelerator

DSP Farm

NPUs

FPGAs

$$$$

27

Nasdaq: MSPD

Sys

tem

s

Hardware Defect Density

Software Defect Density

Hardware

Software

Increasing SoC Complexity …

> 20 Million

Lines Of

“Carrier Class”

Code

Capital Intensive Solution

Labor Intensive Solution

CriticalEnabler

28

Nasdaq: MSPD 29

Increasing SoC Complexity …

Software

Complexity

lines of

code

Silicon

Complexity

number of

transistors

Dial-Up Modems VoIP Multi-RAT Baseband

1995 2005 2010

20

million

> 900

million300

million

< 3 million

< 10,000

300x

2000x

2 million

Nasdaq: MSPD

Mindspeed推荐方案:带天线小基站 (Small Cell Solution With Local Radio)


Up to 102.4Gbps2DDR: 56.2Gbps

4DDR: 102.4Gbps

30

Transcede

SoC

Nasdaq: MSPD

Mindspeed推荐方案:接RRU小基站 (SmallcellSolution Supporting RRU)


Remote RF

Up to 102.4Gbps2DDR: 56.2Gbps

4DDR: 102.4Gbps

31

Transcede

SoC

Nasdaq: MSPD

Enterprise/Home System – LTE/3G + WiFi

Transcede

SoC

Ethernet Switch

RadiSys EPC

Cabled RF

USB

Client laptopDL: PLAY VideoUL: SEND Video

Server laptop

DL: SEND VideoUL: PLAY Video

MindspeedSmall CellReference Unit

HTCSmartphone

Nasdaq: MSPD

Mindspeed推荐方案的特性与优点 (Advantages of Mindspeed Solutions)

Mindspeed的基带系列产品(低端、中端及高端产品)能带来许多前所未有的灵活度并支持多模的各种应用 (Transcede SoC are ideal for a family of SDR solutions capable to support multi-RAT):

- 低端的Transcede芯片可用来实现家庭、或小型企业级Femto基站 (Low end Transcede can support Home or Small Enterprise Femtocell)

• 1x20MHz 2x2 FDD/TDD 或(or) 1x5MHz W-CDMA carrier 或

• 1x15MHz 2x2 FDD/TDD + 1x5MHz W-CDMA 或

• 通过软件配制或升级，实现从W-CDMA到LTE的平滑过渡、升级 (SW upgrade to evolve from W-CDMA to LTE)

- 中端的Transcede芯片可用来实现企业级Pico基站 (Mid-range Transcedesupports more capable Enterprise Pico/Femto)

• 1x20MHz 4x4 FDD/TDD 或(or) 2x20MHz 2x2 FDD/TDD 或(or)

• 2x5MHz W-CDMA载扇或

• 1x5MHz W-CDMA + 1x20MHz 2x2 FDD/TDD多模 (Concurrent Multi-mode)

• 支持LTE TDD、FDD回传+LTE FDD接入 (LTE Backhaul + LTE Access)

• 通过软件配制，实现从多载W-CDMA到不同LTE的平滑过渡、升级(SW upgrade to evolve from W-CDMA to LTE)

- 高端的Transcede芯片能支持更多不同的应用(High end Transcede can support even more capable applications)


Nasdaq: MSPD

现有设备组网状况：宏站＋室分天线 (Current Network Topology: Macrocell+DAS)


BBU 资源池

RRU

RRU

DAS

ＩＰＣｏｒｅ

RRU

ｅＮＢ

ＣＰＲＩ

L1/L2/L3

L1/L2/L3

千兆 (Gb/s) I/Q百兆空口流量(100s Mb/s at

air-interface)

百兆空口流量(100s Mb/s at

air-interface)


air-interface)

Nasdaq: MSPD

基于以上推荐的小基站方案的组网 (Better Network Topology With Multi-RAT Small Cell)


数据汇聚及管理中心

ＩＰＣｏｒｅ

RRU+微基站

L1/L2/L3

In-door Femtocell

FTTx

Out-door PicoHome Femto

FTTx

千兆空口流量(Gb/s at air-interface)

L1/L2/L3

L1/L2/L3

IP/Ethernet

RRU+微基站

L1/L2/L3

RRU+微基站

L1/L2/L3

室外小基站

千兆 (Gb/s) ＩＰ

复用现有光纤资源(Reuse current fibers)

千兆空口流量(Gb/s at air-interface)


air-interface)

Nasdaq: MSPD

总结 (Conclusions)

随着无线移动数据业务的不断发展，LTE的商用势在必然(Wireless Broadband applications will depend on LTE)

- 支持多模同时也就变成不可回避的问题 (Support of Multi-RAT becomes necessary)

高频频段的使用，加上用户对吞吐率需求的不断提高，覆盖扇区小型化的趋势也不可逆转 (Use of high frequency bands, together with ever increasing throughput demand from users, further cell-sectorizationis inevitable)

Mindspeed推荐的多模小基站解决方案能有效地解决目前面对的难题(Transcede based small cell solutions can efficiently resolve multiple problems)

- 室内覆盖、补盲、吞吐密度、CAPEX/OPEX、等等 (coverage, throughput, cost, etc.)

我们希望和业界领先一起来推动这个产业的健康发展，为无线移动宽带业务做贡献 (We welcome collaborations with industry pioneers and our customers to promote a healthy development of wireless broadband services)


Thank You!

Q & A

Multi-RAT Small Cells: An Efficient Solution for 3G/LTE Deployment

Documents