📚 Core Concepts of Pre-U CIE Computer Science | Pre-U CIE 计算机科学核心知识点梳理
The Cambridge Pre-U Computer Science syllabus demands a deep understanding of both theoretical principles and practical application. This revision guide summarises the core knowledge areas, from hardware fundamentals to ethical computing, ensuring you are thoroughly prepared for examination. Each section is presented with concise English explanations immediately followed by Chinese translations, helping you master terminology and concepts bilingually.
剑桥 Pre-U 计算机科学课程要求深入理解理论原理和实际应用。这份复习指南总结了从硬件基础到计算机伦理的核心知识领域，确保你为考试做好充分准备。每个部分先提供简洁的英文解释，紧接着是中文翻译，帮助你双语掌握术语和概念。
1. Data Representation | 数据表示
All data in a computer is stored as binary digits (bits). Unsigned integers use pure binary; signed integers employ two's complement, where the most significant bit indicates sign. For an n-bit number, the range is -2ⁿ⁻¹ to 2ⁿ⁻¹–1. Overflow occurs when an arithmetic result exceeds this range.
计算机中的所有数据都以二进制位（比特）存储。无符号整数使用纯二进制表示；有符号整数采用补码表示，最高位表示符号位。对于n位数字，范围是-2ⁿ⁻¹到2ⁿ⁻¹–1。当算术结果超出此范围时会发生溢出。
Floating-point numbers follow the IEEE 754 standard, splitting a value into sign, mantissa, and exponent. This provides a trade‑off between range and precision. Normalisation ensures a unique representation by adjusting the mantissa so its first digit is always 1.
浮点数遵循IEEE 754标准，将数值分为符号、尾数和指数。这提供了范围和精度之间的权衡。规范化通过调整尾数使第一位始终为1，确保表示形式的唯一性。
Characters are encoded with ASCII (7‑bit, 128 characters) or Unicode (up to 32 bits) to represent global scripts. Images use pixel arrays and colour depth; sound samples are digitised via sampling rate and bit depth.
字符使用ASCII（7位，128个字符）或Unicode（最多32位）进行编码以表示全球文字。图像使用像素阵列和色彩深度；声音样本通过采样率和采样深度被数字化。
2. Computer Architecture | 计算机体系结构
Modern computers are built on the Von Neumann architecture, which stores both program instructions and data in the same memory. The CPU consists of the Control Unit (CU), Arithmetic Logic Unit (ALU), and registers such as the Program Counter (PC), Memory Address Register (MAR), and Memory Data Register (MDR).
现代计算机建立在冯·诺依曼架构之上，将程序指令和数据存储在同一内存中。CPU由控制单元（CU）、算术逻辑单元（ALU）以及程序计数器（PC）、内存地址寄存器（MAR）和内存数据寄存器（MDR）等寄存器组成。
The fetch‑decode‑execute cycle continuously retrieves an instruction, decodes it, and executes it. Buses (address, data, control) carry information between components. Factors affecting performance include clock speed, number of cores, cache size, and word length.
取指－解码－执行循环持续地取出指令、解码并执行。总线（地址、数据、控制）在各组件间传递信息。影响性能的因素包括时钟速度、核心数量、缓存大小和字长。
Memory hierarchy balances cost and speed: registers (fastest), cache (SRAM), main memory (DRAM), and secondary storage (HDD, SSD). Virtual memory uses disk space to simulate larger RAM, managed by the OS through paging.
内存层次结构平衡了成本与速度：寄存器（最快）、缓存（SRAM）、主存（DRAM）和辅助存储（HDD、SSD）。虚拟内存使用磁盘空间模拟更大的RAM，由操作系统通过分页进行管理。
3. Operating Systems and Interrupts | 操作系统与中断
An operating system (OS) manages hardware resources, provides a user interface, handles file management, and ensures security. Key functions include process scheduling, memory management, and handling interrupts. An interrupt is a signal that causes the CPU to pause its current task and execute an Interrupt Service Routine (ISR).
操作系统（OS）管理硬件资源，提供用户界面，处理文件管理并确保安全。关键功能包括进程调度、内存管理和中断处理。中断是一个信号，使CPU暂停当前任务并执行中断服务程序（ISR）。
Interrupts can be hardware‑generated (e.g., I/O completion, timer) or software‑generated (traps/exceptions). The CPU preserves the state of the current process on the stack before executing the ISR. Prioritisation ensures urgent events are serviced promptly.
中断可以是硬件生成的（例如I/O完成、定时器）或软件生成的（陷阱/异常）。CPU在执行ISR之前将当前进程的状态保存在堆栈上。优先级确保紧急事件得到及时服务。
Process scheduling algorithms (Round Robin, Shortest Job First, Priority‑based) allocate CPU time. Memory management uses paging and segmentation to map logical addresses to physical frames, enabling multitasking.
进程调度算法（轮转调度、最短作业优先、基于优先级）分配CPU时间。内存管理使用分页和分段将逻辑地址映射到物理帧，从而实现多任务处理。
4. Networking and Communications | 网络与通信
Computer networks are modelled by layered protocols. The TCP/IP model includes Application, Transport, Internet, and Network Access layers. The OSI model adds more granularity with seven layers. Packet switching breaks data into packets routed independently across the network.
计算机网络由分层协议建模。TCP/IP模型包括应用层、传输层、互联网层和网络接入层。OSI模型通过七层结构增加了更细的粒度。分组交换将数据分成独立路由穿越网络的数据包。
IP addresses (IPv4 32‑bit, IPv6 128‑bit) uniquely identify devices. Subnetting divides a network into smaller segments. Routers forward packets using routing tables. Protocols such as HTTP, FTP, SMTP, and DNS operate at the application layer; TCP ensures reliable, ordered delivery, while UDP offers low‑latency transmission.
IP地址（IPv4 32位，IPv6 128位）唯一标识设备。子网划分将网络分割为更小的部分。路由器使用路由表转发数据包。HTTP、FTP、SMTP和DNS等协议在应用层运行；TCP确保可靠有序的交付，而UDP提供低延迟传输。
Network security is enhanced by firewalls (packet filtering, stateful inspection) and encryption (SSL/TLS). Common topologies include star, mesh, and bus, each with trade‑offs in cost, reliability, and scalability.
防火墙（包过滤、状态检测）和加密（SSL/TLS）增强了网络安全性。常见的拓扑结构包括星型、网状和总线型，每种在成本、可靠性和可扩展性上都有不同的权衡。
5. Databases and SQL | 数据库与SQL
A relational database organises data into tables (relations) with rows (tuples) and columns (attributes). Primary keys uniquely identify each row; foreign keys establish relationships between tables. Normalisation reduces data redundancy and prevents update anomalies.
关系数据库将数据组织成表（关系），包含行（元组）和列（属性）。主键唯一标识每一行；外键在表之间建立关系。规范化减少数据冗余并防止更新异常。
First Normal Form (1NF) requires atomic values; 2NF removes partial dependencies on a composite key; 3NF eliminates transitive dependencies. Structured Query Language (SQL) is used to define (DDL), manipulate (DML), and control (DCL) data. Essential commands include SELECT, INSERT, UPDATE, DELETE, and JOIN.
第一范式（1NF）要求原子值；第二范式（2NF）消除了对复合键的部分依赖；第三范式（3NF）消除了传递依赖。结构化查询语言（SQL）用于定义（DDL）、操作（DML）和控制（DCL）数据。基本命令包括SELECT、INSERT、UPDATE、DELETE和JOIN。
Database management systems (DBMS) provide concurrent access, security, backup, and recovery. ACID properties (Atomicity, Consistency, Isolation, Durability) guarantee reliable transaction processing.
数据库管理系统（DBMS）提供并发访问、安全、备份和恢复。ACID特性（原子性、一致性、隔离性、持久性）保证了可靠的事务处理。
6. Algorithm Design and Efficiency | 算法设计与效率
Algorithms solve problems through a finite sequence of unambiguous steps. Common search algorithms include linear search (O(n)) and binary search (O(log n)) on sorted data. Sorting algorithms like bubble sort, insertion sort, merge sort, and quicksort illustrate different time and space complexities.
算法通过有限的无歧义步骤序列解决问题。常见的搜索算法包括线性搜索（O(n)）和对已排序数据的二分搜索（O(log n)）。冒泡排序、插入排序、归并排序和快速排序等排序算法展示了不同的时间和空间复杂度。
Big‑O notation expresses the upper bound of an algorithm's growth rate. Efficient algorithms are vital for large datasets. Recursion is a technique where a function calls itself with a smaller instance, underpinning divide‑and‑conquer strategies.
大O表示法表示算法增长率的上界。对于大型数据集，高效的算法至关重要。递归是一种函数用更小的实例调用自身的技术，是分治策略的基础。
Abstract Data Types (ADTs) such as stacks, queues, linked lists, trees, and graphs model data organisation. Each ADT supports a set of operations (push, pop, enqueue, dequeue) and can be implemented using arrays or pointers.
抽象数据类型（ADT），如栈、队列、链表、树和图，对数据组织进行建模。每种ADT支持一组操作（push、pop、enqueue、dequeue），并且可以使用数组或指针来实现。
7. Programming Concepts and Paradigms | 编程概念与范式
Programming paradigms define the style of coding. Imperative programming uses statements that change program state (e.g., Python, C). Object‑oriented programming (OOP) organises code into classes and objects, emphasising encapsulation, inheritance, and polymorphism.
编程范式定义了编码风格。命令式编程使用改变程序状态的语句（例如Python、C）。面向对象编程（OOP）将代码组织为类和对象，强调封装、继承和多态。
Declarative paradigms (functional, logic) focus on what the program should accomplish without specifying control flow. Functional programming treats computation as evaluation of mathematical functions, avoiding side effects. Logic programming (e.g., Prolog) uses facts and rules to deduce answers.
声明式范式（函数式、逻辑式）侧重于程序应完成什么，而不指定控制流程。函数式编程将计算视为数学函数的求值，避免副作用。逻辑编程（例如Prolog）使用事实和规则来推导答案。
Source code is translated by compilers (entire program to machine code) or interpreters (line‑by‑line). Linking and loading combine object files into an executable. Parameter passing mechanisms include pass‑by‑value, pass‑by‑reference, and pass‑by‑name.
源代码通过编译器（整个程序转成机器码）或解释器（逐行执行）进行翻译。链接和加载将目标文件合并为可执行文件。参数传递机制包括按值传递、按引用传递和按名传递。
8. System Development Life Cycle | 系统开发生命周期
The System Development Life Cycle (SDLC) provides a structured approach to creating information systems. Traditional models include the Waterfall model (linear sequential phases) and the Spiral model (iterative risk‑driven). Agile methodologies, such as Scrum, emphasise incremental delivery and rapid feedback.
系统开发生命周期（SDLC）为创建信息系统提供了一种结构化方法。传统模型包括瀑布模型（线性顺序阶段）和螺旋模型（迭代风险驱动）。敏捷方法，如Scrum，强调增量交付和快速反馈。
Feasibility study evaluates technical, economic, legal, and operational viability. Requirements analysis produces a specification document using techniques like interviews and prototyping. Design translates requirements into system architecture, data models, and user interfaces.
可行性研究评估技术、经济、法律和运营的可行性。需求分析使用访谈和原型设计等技术生成规格文档。设计将需求转化为系统架构、数据模型和用户界面。
Testing verifies that the system meets its specification. Levels include unit testing, integration testing, system testing, and acceptance testing. Maintenance after deployment includes corrective, adaptive, and perfective activities.
测试验证系统是否满足其规格。测试级别包括单元测试、集成测试、系统测试和验收测试。部署后的维护包括纠正性、适应性和完善性活动。
9. Boolean Algebra and Digital Logic | 布尔代数与数字逻辑
Boolean algebra manipulates logical variables with operators AND ( · ), OR (+), and NOT (¬). Boolean expressions can be simplified using laws such as De Morgan's theorems: ¬(A · B) = ¬A + ¬B and ¬(A + B) = ¬A · ¬B.
布尔代数使用AND（ · ）、OR（+）和NOT（¬）运算符处理逻辑变量。布尔表达式可以使用德摩根定律进行化简：¬(A · B) = ¬A + ¬B 和 ¬(A + B) = ¬A · ¬B。
F = A · ¬B + ¬A · B (XOR)
| A | B | A XOR B |
| 0 | 0 | 0 |
| 0 | 1 | 1 |
| 1 | 0 | 1 |
| 1 | 1 | 0 |
Logic gates implement Boolean functions electronically. Combining NAND gates can produce any other gate (functional completeness). Half‑adders and full‑adders form the basis of arithmetic circuits; a full adder handles carry‑in, producing sum and carry‑out.
逻辑门通过电子方式实现布尔函数。结合NAND门可以生成任何其他门（功能完备性）。半加器和全加器构成了算术电路的基础；全加器处理进位输入，产生和与进位输出。
Karnaugh maps (K‑maps) provide a visual method to simplify Boolean expressions up to 6 variables, helping minimise the number of logic gates in circuit design. Flip‑flops (SR, JK, D) are sequential circuits that store a single bit and form registers and memory cells.
卡诺图（K‑maps）提供了一种可视化方法，可简化最多6个变量的布尔表达式，有助于减少电路设计中的逻辑门数量。触发器（SR、JK、D）是存储单个位的时序电路，构成寄存器和存储单元。
10. Information Security and Cryptography | 信息安全与密码学
Information security aims to protect confidentiality, integrity, and availability (CIA triad). Threats include malware, phishing, denial‑of‑service (DoS), and man‑in‑the‑middle attacks. Countermeasures involve authentication, authorisation, and encryption.
信息安全旨在保护机密性、完整性和可用性（CIA三元组）。威胁包括恶意软件、网络钓鱼、拒绝服务（DoS）和中间人攻击。对策包括身份验证、授权和加密。
Symmetric encryption uses a single shared key (e.g., AES, DES). Asymmetric encryption employs a key pair – public and private – enabling secure key exchange and digital signatures (e.g., RSA). Hashing functions (MD5, SHA‑256) produce a fixed‑length digest, used to verify data integrity.
对称加密使用单个共享密钥（例如AES、DES）。非对称加密使用一对密钥——公钥和私钥——实现安全密钥交换和数字签名（例如RSA）。哈希函数（MD5、SHA‑256）生成固定长度的摘要，用于验证数据完整性。
Digital certificates bind a public key to an entity via a trusted Certificate Authority (CA). Firewalls, intrusion detection systems (IDS), and regular patches provide defence layers. The principle of least privilege limits user access rights.
数字证书通过可信的证书颁发机构（CA）将公钥绑定到实体。防火墙、入侵检测系统（IDS）和定期补丁提供了防御层。最小权限原则限制了用户的访问权限。
11. Computational Thinking and Turing Machines | 计算思维与图灵机
Computational thinking involves abstraction, decomposition, pattern recognition, and algorithmic design. It underpins problem‑solving across all computer science. Finite state machines (FSMs) model systems with a finite number of states, inputs, and transitions; they are used in protocol design and lexical analysis.
计算思维包括抽象、分解、模式识别和算法设计。它是整个计算机科学领域解决问题的基础。有限状态机（FSM）用有限数量的状态、输入和转换对系统建模；它们被应用于协议设计和词法分析。
A Turing machine is a theoretical model consisting of an infinite tape, a read/write head, a set of states, and transition rules. It can simulate any algorithmic process, defining the limits of computability. The Halting Problem proves that some problems are undecidable.
图灵机是一种理论模型，由无限长的带子、读写头、一组状态和转换规则组成。它可以模拟任何算法过程，定义了可计算性的界限。停机问题证明有些问题是不可判定的。
Regular expressions and context‑free grammars (Backus‑Naur Form) describe the syntax of programming languages. The Chomsky hierarchy classifies languages from regular to recursively enumerable, linking theory to compiler and interpreter design.
正则表达式和上下文无关文法（巴科斯‑诺尔范式）描述编程语言的语法。乔姆斯基层次结构将语言从正则语法分类到递归可枚举语法，将理论与编译器、解释器设计联系起来。
12. Ethical, Legal, and Professional Issues | 伦理、法律与职业问题
Computing professionals must adhere to codes of conduct, such as those from the BCS or ACM, emphasising public interest, integrity, and competence. Data protection laws (e.g., GDPR) regulate the collection, storage, and processing of personal data, requiring consent and the right to erasure.
计算机专业人士必须遵守行为准则，如BCS或ACM的准则，强调公共利益、诚信和能力。数据保护法（例如GDPR）规范了个人数据的收集、存储和处理，要求取得同意并赋予删除权。
Intellectual property rights cover software licensing (proprietary, open source, freeware) and copyright. Plagiarism and unauthorised copying are ethical and legal violations. The digital divide refers to inequalities in access to technology, raising social and ethical concerns.
知识产权涵盖软件许可（专有、开源、免费软件）和版权。剽窃和未经授权的复制是违反伦理和法律的行为。数字鸿沟指的是在技术获取方面的不平等，引发了社会和伦理问题。
Environmental impacts of computing include energy consumption of data centres and e‑waste. Artificial intelligence raises questions about bias, accountability, and job displacement. Students should be able to analyse case studies, balancing innovation with responsibility.
计算机对环境的影响包括数据中心的能源消耗和电子垃圾。人工智能引发了关于偏见、问责制和工作岗位替代的问题。学生应能够分析案例研究，在创新与责任之间取得平衡。
Published by TutorHao | Computer Science Revision Series | aleveler.com
更多咨询请联系16621398022（同微信）
或截图后用「扫一扫 → 相册」选择本图