網(wǎng)絡(luò)處理器通常由多個(gè)異構(gòu)的處理和內(nèi)存單元通過(guò)片上網(wǎng)絡(luò)連接構(gòu)成，其目標(biāo)應(yīng)用需要在Gbit/s到幾十Gbit/s 的網(wǎng)絡(luò)環(huán)境中以線速處理數(shù)據(jù)包?；诰W(wǎng)絡(luò)處理器的應(yīng)用有實(shí)時(shí)、資源受限和異構(gòu)的特點(diǎn)。組件技術(shù)對(duì)于復(fù)雜的嵌入式系統(tǒng)是一種十分有前途的方法。本文以一種典型的網(wǎng)絡(luò)處理器為例，說(shuō)明了在基于網(wǎng)絡(luò)處理器的系統(tǒng)中應(yīng)用組件技術(shù)時(shí)，對(duì)組件框架的要求，討論了組件組合框架和運(yùn)行時(shí)框架，并定義了組件框架服務(wù)。利用提出的組件框架，可以實(shí)現(xiàn)軟件性能工程。
關(guān)鍵詞: 網(wǎng)絡(luò)處理器; 組件模式; 組件框架; 軟件性能工程.

ABSTRACT: Network processor based applications are real-time, resource-constrained and heterogeneous. We believe that component technology is a promising approach for complex embedded system development. However, the general purpose component technology is not suitable for the contexts of network processor. This paper gives requirements of component framework for applying component technology to network processor based systems.
Composition and run-time framework are discussed. The proposed component framework supports software performance engineering.
KEY WORDS: Network processor, Component model, Component framework, Software performance engineering