卫星通信系统中的跨层动态带宽分配优化算法

需积分: 25 104 浏览量更新于2024-09-15 收藏 1.47MB PDF 举报

"该文提出了一种用于优化数字视频广播(DVB)-回程卫星(RCS)系统动态带宽分配(DBA)的新型跨层框架，采用自适应编码。通过在网络子网内应用网络效用最大化(NUM)框架，将考虑自适应物理层和高层服务质量(QoS)需求的媒体访问控制(MAC)方法设计为优化问题。文中提出了符合DVB-RCS标准的分层和全局解决策略，不仅提供最小带宽保证，还强调公平性。同时，它们允许在最小化信令开销的同时，对时隙大小和整体系统效率进行联合优化。此外，还介绍了一种解决DBA问题的低计算复杂度算法，实际操作中能增加具有绝对和相对QoS要求的连接数量，并促进卫星网络在互联网协议(IP)环境中的互操作性。" 本文重点探讨了雷达湍流信号检测算法在卫星通信系统中的应用。DVB-RCS系统是一种先进的卫星通信技术，旨在提供高数据速率的双向服务，如宽带互联网接入。在这样的系统中，动态带宽分配是关键，因为它直接影响到服务质量、系统效率和用户公平性。文章中提出的跨层框架打破了传统的网络设计范式，将物理层与高层协议（如MAC层和QoS）相结合，通过网络效用最大化来优化带宽分配。这种框架利用自适应编码，可以根据信道条件的变化实时调整编码率，从而有效应对雷达湍流等引起的信道波动，确保传输的可靠性。在优化问题的求解上，文章提出了一种层次化和全局化的策略，这既满足了DVB-RCS标准，又实现了带宽保证和公平性的最大化。此外，通过联合优化时隙大小和整体系统效率，能够在减少信令开销的同时提高系统的运行效率。所提出的低计算复杂度算法降低了处理DBA问题的复杂性，使得系统能够处理更多的连接，且能更好地适应具有特定QoS需求的不同用户。在实际应用中，这一算法的实施能够显著增强DVB-RCS系统处理雷达湍流信号的能力，提升在恶劣环境下的通信性能。同时，由于其考虑了IP环境的兼容性，有助于实现卫星网络与其他IP网络的无缝对接，促进了卫星通信系统的互操作性和广泛应用。

64 IEEE SYSTEMS JOURNAL, VOL. 2, NO. 1, MARCH 2008

Fig. 2. Scheduling (bandwidth allocation) problem.

NP-hard (not solvable in polynomial time). It is actually mean-

ingful to impose some structure on the MF-TDMA and in this

section we deal with those issues, whereas the allocation itself

is left to the following section.

In DVB-RCS, the allocation of resources is a reaction to the

capacity requests performed by the RCSTs (or users, equiva-

lently). As deﬁned in the standard, the TBTP shall be updated

and transmitted every SF, whereas bandwidth is allocated at a

frame level. Let us assume that the SF is divided into

frames

using the whole SF bandwidth

and with a time du-

ration

, where is the SF duration (typically

265 ms). We further assume that

is divided into dif-

ferent carrier types to accommodate different users accounting

for different service level agreements (SLAs), terminal equip-

ment, or location, so that an RCST uses only one type of carrier.

Under these assumptions, the global allocation is decoupled into

independent sub-allocations ( standing for the number of

carrier types). [See Fig. 2 (left side).]

Thus, the problem we consider consists in multiplexing

users into carriers of bandwidth that transmit during

seconds (see Fig. 2, right side). Without loss of performance and

to facilitate upcoming issues, we group all RCSTs that transmit

within the same carrier type (equivalently symbol rate) and the

same coding rate (in the DVB-RCS, adaptive coding is envis-

aged to compensate the physical quality of the transmission, i.e.,

channel conditions). In accordance with ETSI technical speciﬁ-

cation [19], we refer to each of those groups as an area (we can

interpret areas as the earth surface zones where channel condi-

tions are similar).

We consider bandwidth allocation (sometimes referred as

scheduling) at MAC layer. In the rest of this paper, we talk

about scheduling ATM cells (53 bytes), but the method is valid

for any packet length. As an example, the DVB-RCS standard

describes moving picture experts group (MPEG) containers

(188 bytes) as a possible MAC-layer unit.

We introduce now the key aspect of the proposed framework,

which establishes the tradeoff choice between complexity and

optimality. A TS of duration

is ﬁxed common to all areas

in each allocation process. In Fig. 2, the idea is depicted with

Fig. 3. Bandwidth allocation hierarchy in DVB-RCS.

only two areas. Due to the different coding rates, the transmis-

sion time of an ATM cell varies from area to area and so varies

the percentage of the time the TS is used, i.e., the bandwidth ef-

ﬁciency per area. Note that it is possible to transmit more than

one ATM cell per TS and more speciﬁcally, the DVB-RCS takes

into account 1, 2, or 4 ATM cells per TS. In this way, cross-layer

information from the PHY layer is taken into account. To get

cross-layer information from the upper layers, we propose to

use the 4 bits available in the ﬁeld

Channel_ID available at SAC

messages. Note that this ﬁeld remains unused if the satellite is

transparent. In this way, it is possible to distinguish different

trafﬁc types that request capacity using the same type of ca-

pacity request. For example, we can consider the QoS deﬁned

at IP-level in order to conﬁgure a satellite subnetwork as trans-

parent as possible at TCP level.

Further issues of the architecture are both a reduction in

signalling and an increased robustness to RCSTs’ PHY-layer

changes. Regarding signalling, note that TSs with the same

characteristics need to be deﬁned only once (with repetitions)

in FCT and TCT tables. The reader can ﬁnd in [17] and [20],

different approaches that consider changing the timeslot dura-

tion as a function of the coding rate of the area. Advantages

and disadvantages of such an approach have already been

introduced in this section. Adding now robustness issues to the

discussion, note that in full-ﬂexible solutions, the PHY-layer

changes in the RCSTs require a whole frame redesign, whereas

this is not the case with a common TS approach.

For the sake of completeness, we include the hierarchical

bandwidth allocation concept deﬁned in [19]. The motivation

is to guarantee some minimum resources to service providers

(SPs) as an extra mechanism to grant QoS to their corresponding

RCSTs. Note that not necessarily all RCSTs attached to a cer-

tain SP see the same channel conditions and thus, they do not

belong to the same area in general. It is then adequate to de-

ﬁne the segment concept as the group of users that belong to an

SP and to guarantee some satellite resources to each segment

(e.g., depending on how much they pay). See both area and seg-

ment divisions in Fig. 3 and note that they do not deﬁne disjoint

剩余11页未读，继续阅读

lxy3103

粉丝: 3
资源: 9

卫星通信系统中的跨层动态带宽分配优化算法

心电信号检测算法

MIMO/MASSIVE MIMO 信号检测算法matlab仿真程序

信号检测算法

湍流气象目标的检测算法matlab仿真程序，用于气象雷达信号处理.zip

增强激光雷达湍流廓线反演的改进算法

一种机载合成孔径成像激光雷达聚束模式成像算法

main_雷达回波数据_雷达杂波_雷达CLEAN算法_雷达杂波仿真_雷达_源码.zip

"基于卷积神经网络的激光雷达湍流探测与预警研究

机载合成孔径激光雷达聚束模式成像算法研究

雷达信号处理中的自适应滤波技术：智能算法的较量

最新资源