A-Level AQA Computer Science: Last-Minute Revision Notes | A-Level AQA 计算机：考前冲刺笔记

📚 A-Level AQA Computer Science: Last-Minute Revision Notes | A-Level AQA 计算机：考前冲刺笔记

This revision guide condenses the entire AQA A-Level Computer Science specification into concise, exam-focused notes. Use it to consolidate your understanding of key concepts, from programming paradigms to computer architecture, algorithms, and the legal framework surrounding technology. Each section follows a mirrored bilingual format to strengthen recall and clarity.

这份复习指南将 AQA A-Level 计算机科学全部考纲浓缩为精炼的考点笔记。用它来巩固你对核心概念的理解，内容涵盖编程范式、计算机体系结构、算法以及技术相关法律框架。每个部分采用中英对照的格式，帮助你强化记忆和厘清思路。

1. Fundamentals of Programming | 编程基础

A variable is a named memory location whose value can change during execution, while a constant retains a fixed value. Data types include integer, real/float, Boolean, character, and string. Always declare variables with an explicit type in high-level languages such as Pascal or Visual Basic, though Python uses dynamic typing.

变量是命名的内存位置，其值在执行过程中可以改变；常量则保持固定值。数据类型包括整型、实型/浮点型、布尔型、字符型和字符串。在 Pascal 或 Visual Basic 等高级语言中要显式声明类型，而 Python 使用动态类型。

Selection (IF…THEN…ELSE, CASE/SWITCH) and iteration (FOR, WHILE, REPEAT…UNTIL) form the backbone of structured programs. Nested structures allow complex decision-making. Subroutines (procedures and functions) are reusable blocks; parameters can be passed by value (a local copy) or by reference (the original variable). A function returns a single value, whereas a procedure does not.

选择结构 (IF…THEN…ELSE, CASE/SWITCH) 和迭代结构 (FOR, WHILE, REPEAT…UNTIL) 是结构化程序的骨干。嵌套结构实现复杂决策。子程序（过程和函数）是可重用的代码块；参数可以通过传值（局部副本）或传引用（原始变量）传递。函数返回一个值，过程则不返回。

Recursion is a technique where a subroutine calls itself. It requires a stopping condition (base case) to prevent infinite recursion. Recursion often provides elegant solutions for tree traversal, factorial calculation, and the Towers of Hanoi, but can be less memory-efficient due to call stack usage.

递归是一种子程序调用自身的技术。它需要一个停止条件（基准情形）以避免无限递归。递归常为树的遍历、阶乘和汉诺塔问题提供优雅的解法，但因调用栈的开销可能内存效率较低。

2. Data Structures | 数据结构

Arrays are collections of elements of the same data type, stored in contiguous memory locations. They can be one-dimensional (list), two-dimensional (table), or multi-dimensional. Static arrays have a fixed size declared at compile time; dynamic arrays can be resized at runtime, often using heap memory.

数组是同一数据类型元素的集合，存储在连续内存位置。可以是一维（列表）、二维（表）或多维。静态数组在编译时声明固定大小；动态数组可在运行时调整大小，通常使用堆内存。

Abstract data types (ADTs) like stacks, queues, and priority queues can be implemented using arrays or linked lists. A stack is LIFO (Last In First Out) with operations push and pop; a queue is FIFO (First In First Out) with enqueue and dequeue. A linked list stores items non-contiguously, with each node containing data and a pointer to the next node.

抽象数据类型（ADT），如栈、队列和优先级队列，可以用数组或链表实现。栈是后进先出（LIFO），操作有压入和弹出；队列是先进先出（FIFO），有入队和出队操作。链表非连续存储元素，每个节点包含数据和指向下一节点的指针。

Graphs are represented by vertices (nodes) and edges (arcs). An adjacency matrix uses a 2D array to show connections, while an adjacency list stores a list of neighbours for each vertex. Trees are hierarchical graphs where each node has one parent (except root) and zero or more children. Binary trees have at most two children, used in binary search trees and expression trees.

图由顶点（节点）和边（弧）表示。邻接矩阵用二维数组表示连接，邻接表则为每个顶点存储相邻顶点的列表。树是层次结构的图，每个节点（除根节点外）有一个父节点和零或多个子节点。二叉树最多有两个子节点，用于二叉搜索树和表达式树。

3. Algorithms | 算法

Searching algorithms: linear search checks each element sequentially (O(n)), suitable for unsorted data. Binary search repeatedly divides a sorted dataset in half (O(log n)), much faster on large data. Sorting algorithms: bubble sort compares adjacent pairs and swaps (O(n²)), insertion sort builds a sorted sublist (O(n²)), merge sort uses divide-and-conquer (O(n log n)) and is stable.

搜索算法：线性搜索顺序检查每个元素 (O(n))，适用于未排序数据。二分搜索不断将有序数据集对半分 (O(log n))，处理大数据时快得多。排序算法：冒泡排序比较相邻对并交换 (O(n²))，插入排序构建已排序子列表 (O(n²))，归并排序采用分治策略 (O(n log n)) 且是稳定排序。

Algorithm efficiency is measured using Big-O notation, which describes the upper bound of time or space complexity as input size n grows. Constant O(1), logarithmic O(log n), linear O(n), linearithmic O(n log n), quadratic O(n²), and exponential O(2ⁿ) are common orders. Understanding complexity helps select appropriate algorithms for large datasets.

算法效率用大O记号衡量，描述随着输入规模 n 增长的时间或空间复杂度上界。常见阶有常数 O(1)，对数 O(log n)，线性 O(n)，线性对数 O(n log n)，平方 O(n²) 和指数 O(2ⁿ)。理解复杂度有助于为大数据集选择合适的算法。

Graph traversal: depth-first search (DFS) uses a stack to go as deep as possible before backtracking. Breadth-first search (BFS) uses a queue to explore neighbours level by level. Dijkstra’s algorithm finds the shortest path in a weighted graph with non-negative weights, using a priority queue.

图的遍历：深度优先搜索 (DFS) 使用栈尽可能深入再回溯。广度优先搜索 (BFS) 使用队列逐层遍历邻居。迪杰斯特拉算法使用优先级队列找出带非负权值图的单源最短路径。

4. Theory of Computation | 计算理论

Finite State Machines (FSMs) model systems with a finite number of states, inputs, and transitions. They can be represented as state transition diagrams or tables. A Mealy machine outputs depend on state and input, while a Moore machine’s outputs depend only on state. FSM can be used for recognisers, vending machines, and protocol design.

有限状态机 (FSM) 用有限的状态、输入和转移对系统建模，可用状态转移图或表表示。米利型机器的输出取决于状态和输入，摩尔型机器的输出仅取决于状态。FSM 用于识别器、自动售货机和协议设计。

Regular expressions describe patterns in strings. Symbols include: * (zero or more), + (one or more), ? (zero or one), | (or), . (any character). They are used in lexical analysis and input validation. A language is regular if it can be expressed by a finite automaton; certain languages (e.g., balanced brackets) require context-free grammars.

正则表达式描述字符串模式。符号包括：*（零或多），+（一或多），?（零或一），|（或），.（任意字符）。用于词法分析和输入验证。如果一个语言可以被有限自动机表示则是正则语言；某些语言（如匹配括号）需要上下文无关文法。

Turing machines are abstract models of computation: an infinite tape, a head reading/writing symbols, a finite set of states and a transition function. They can simulate any algorithm and define the limits of computation. The Halting Problem shows that no algorithm can decide whether an arbitrary program terminates — a fundamental undecidable problem.

图灵机是计算抽象模型：无限长纸带、读写符号的磁头、有限状态和转移函数。它能模拟任何算法并定义了计算的极限。停机问题表明没有算法能够判定任意程序是否终止——这是一个基本的不可判定问题。

5. Computer Systems | 计算机系统

A computer system consists of hardware, software (system and application), and data. The operating system (OS) manages resources (scheduling, memory, I/O) and provides a user interface. Virtual memory uses disk space as an extension of RAM, swapping pages in and out. Paging and segmentation are memory management techniques.

计算机系统由硬件、软件（系统和应用）和数据组成。操作系统 (OS) 管理资源（调度、内存、I/O）并提供用户接口。虚拟内存使用磁盘空间扩展 RAM，以页面换入换出。分页和分段是内存管理技术。

System software includes translators: assembler (assembly to machine code), compiler (high-level language to machine code in one go), and interpreter (translates and executes line by line). Compilers produce intermediate object code and require linking. Bytecode (as in Java) is compiled to an intermediate form run on a virtual machine.

系统软件包括翻译器：汇编器（汇编到机器码），编译器（将高级语言一次性翻译为机器码），解释器（逐行翻译并执行）。编译器生成中间目标代码并需要链接。字节码（如 Java）被编译为中间形式，在虚拟机上运行。

Data representation: Binary (base 2), hexadecimal (base 16), and two’s complement for signed integers. Floating-point numbers are stored as mantissa × baseᵉˣᵖᵒⁿᵉⁿᵗ. Normalisation ensures maximum precision. ASCII uses 7 bits per character; Unicode (e.g., UTF-8) supports a vast range of characters.

数据表示：二进制（基2）、十六进制（基16）以及用补码表示有符号整数。浮点数存储为尾数 × 基数指数。规格化确保最大精度。ASCII 每字符用 7 位；Unicode（如 UTF-8）支持大量字符。

6. Computer Organisation and Architecture | 计算机组成与体系结构

The stored program concept (von Neumann architecture) uses a single memory for both data and instructions. The CPU consists of the Control Unit (CU), Arithmetic Logic Unit (ALU), and registers (PC, MAR, MDR, CIR, accumulator). The fetch-decode-execute cycle is at the heart of instruction processing.

存储程序概念（冯·诺依曼体系结构）使用同一内存存放数据和指令。CPU 由控制单元 (CU)、算术逻辑单元 (ALU) 和寄存器（PC, MAR, MDR, CIR, 累加器）组成。取指-译码-执行周期是指令处理的核心。

Factors affecting processor performance: clock speed (GHz), number of cores, and cache memory size/levels. Pipelining increases throughput by overlapping instruction stages. Parallel processing can be SIMD or MIMD. Harvard architecture separates instruction and data memories, allowing simultaneous access.

影响处理器性能的因素：时钟频率 (GHz)、内核数量以及缓存大小/层级。流水线通过重叠指令阶段提高吞吐量。并行处理可以是 SIMD 或 MIMD。哈佛体系结构将指令和数据内存分开，允许同时访问。

Logic gates (AND, OR, NOT, NAND, NOR, XOR) combine to form combinatorial and sequential circuits. A half adder adds two bits producing sum (S) and carry (C); a full adder also includes a carry-in. D-type flip-flops store a single bit and act as memory elements, forming registers and counters.

逻辑门（与、或、非、与非、或非、异或）组成组合电路和时序电路。半加器将两个比特相加产生和 (S) 与进位 (C)；全加器还包括进位输入。D 型触发器存储单个比特，充当存储元件，构成寄存器和计数器。

Sum = A ⊕ B, Carry = A ∧ B

全加器：S = A ⊕ B ⊕ Cᵢₙ, Cₒᵤₜ = (A ∧ B) ∨ (Cᵢₙ ∧ (A ⊕ B))

7. Fundamentals of Communication and Networking | 通信与网络基础

Network topologies: bus (single backbone, collisions), star (central switch, single point of failure but manageable), mesh (each node connected to several others, redundant). Wi-Fi uses CSMA/CA to avoid collisions. Ethernet LANs use CSMA/CD. The TCP/IP stack comprises application, transport (TCP/UDP), internet (IP), and link layers.

网络拓扑：总线型（单主干，冲突），星型（中央交换机，单点故障但易管理），网状（每个节点连接到多个其他节点，冗余）。Wi-Fi 使用 CSMA/CA 避免冲突。以太局域网使用 CSMA/CD。TCP/IP 协议栈包含应用层、传输层 (TCP/UDP)、网际层 (IP) 和链路层。

IP addressing: IPv4 (32-bit, dotted decimal) and IPv6 (128-bit, hexadecimal). Routers forward packets using IP address and routing tables. NAT (Network Address Translation) allows multiple devices to share one public IP. The Domain Name System (DNS) translates domain names to IP addresses.

IP 地址：IPv4（32位，点分十进制）和 IPv6（128位，十六进制）。路由器使用 IP 地址和路由表转发数据包。NAT（网络地址转换）允许多台设备共享一个公网 IP。域名系统 (DNS) 将域名解析为 IP 地址。

Client-server and peer-to-peer are two network models. In client-server, a central server provides services; in P2P, each node acts as both client and server. Firewalls (packet filtering, stateful inspection) and encryption (symmetric and asymmetric/public-key) are vital for security.

客户端-服务器和对等网络是两种网络模型。客户端-服务器中，中心服务器提供服务；P2P 中，每个节点同时作为客户端和服务器。防火墙（包过滤、状态检测）和加密（对称与非对称/公钥）对安全至关重要。

8. Databases | 数据库

A relational database organises data into tables (relations) linked by primary and foreign keys. Each table has rows (tuples) and columns (attributes). Normalisation (1NF, 2NF, 3NF) reduces data redundancy and update anomalies. First Normal Form requires atomic values; 2NF removes partial key dependencies; 3NF eliminates transitive dependencies.

关系数据库将数据组织为表（关系），由主键和外键连接。每张表有行（元组）和列（属性）。规范化（1NF、2NF、3NF）减少数据冗余和更新异常。第一范式要求原子值；2NF 去除部分键依赖；3NF 消除传递依赖。

SQL (Structured Query Language) is used to define and manipulate data. SELECT ... FROM ... WHERE ... retrieves data; INSERT INTO, UPDATE, DELETE modify records. Joins (INNER, LEFT, RIGHT) combine related tables. Client-server databases offer concurrent access with record locking and transaction management (ACID).

SQL（结构化查询语言）用于定义和操作数据。SELECT ... FROM ... WHERE ... 检索数据；INSERT INTO、UPDATE、DELETE 修改记录。连接（内连接、左连接、右连接）组合相关表。客户端-服务器数据库通过记录锁定和事务管理 (ACID) 提供并发访问。

9. Functional Programming and Big Data | 函数式编程与大数据

Functional programming treats computation as evaluation of mathematical functions, avoiding mutable data and state changes. Core concepts: first-class functions, higher-order functions (map, filter, fold/reduce), and recursion. In Haskell-like syntax, map (+1) [1,2,3] yields [2,3,4]. List comprehension provides a concise way to construct lists.

函数式编程将计算视为数学函数求值，避免可变数据和状态改变。核心概念：一等函数、高阶函数（map、filter、fold/reduce）和递归。在 Haskell 风格语法中，map (+1) [1,2,3] 产生 [2,3,4]。列表推导式提供构建列表的简洁方式。

Big Data refers to datasets too large/complex for traditional processing. The ‘3 Vs’: Volume, Velocity, and Variety. Distributed computing frameworks like Hadoop (MapReduce) and Spark enable parallel processing across clusters. Data mining uncovers patterns, while machine learning builds predictive models.

大数据指过于庞大或复杂以至于传统处理方式无法处理的数据集。“3V”特征：容量 (Volume)、速度 (Velocity) 和多样性 (Variety)。Hadoop (MapReduce) 和 Spark 等分布式计算框架支持跨集群并行处理。数据挖掘发现模式，机器学习构建预测模型。

10. Consequences of Uses of Computing & Exam Tips | 计算机使用的后果与考试技巧

Legal and ethical issues: The Data Protection Act (DPA) governs the use of personal data (fair, lawful, accurate). The Computer Misuse Act makes unauthorised access and hacking illegal. The Regulation of Investigatory Powers Act (RIPA) allows surveillance. Copyright, Designs and Patents Act protects software licensing. Ethical debates cover AI, automation, digital divide, and privacy.

法律与道德议题：《数据保护法》(DPA) 规范个人信息的使用（公平、合法、准确）。《计算机滥用法》将未经授权的访问和黑客行为入罪。《调查权力规管法》(RIPA) 允许监控。《版权、设计和专利法》保护软件许可。伦理辩论涵盖人工智能、自动化、数字鸿沟和隐私。

Environment: e-waste disposal and energy consumption of data centres are sustainability concerns. The circular economy encourages recycling and reducing toxic materials. Exam success: read scenario-based questions carefully; always refer to SPECIFIC sections of the AQA skeleton code pre-released material. Practice tracing algorithms by hand and writing precise pseudocode. Manage time: section A short answers, then section B longer structured questions, finally section C programming.

环境：电子垃圾处理和数据中心的能源消耗是可持续性问题。循环经济鼓励回收和减少有毒材料。考试成功要诀：仔细阅读情景题；始终参照 AQA 预先发布的骨架代码中的具体部分。练习手工追踪算法并编写精确的伪代码。管理时间：先做 A 部分简答题，然后 B 部分较长的结构化问题，最后 C 部分编程题。

Published by TutorHao | Computer Science Revision Series | aleveler.com

更多咨询请联系16621398022（同微信）

A-Level AQA Computer Science: Last-Minute Revision Notes | A-Level AQA 计算机：考前冲刺笔记